Abstract
Objective
Frataxin (FXN) is a mitochondrial protein involved in various cellular processes. To date, the role of FXN in the initiation and progression of colon adenocarcinoma (COAD) remains unclear. This study aimed to investigate FXN expression in COAD and its involvement in COAD initiation and progression.
Methods
FXN expression in COAD was assessed via bioinformatics and immunohistochemistry (IHC) validation. The association between FXN expression and prognosis was evaluated using Kaplan‑Meier analysis and Cox regression analysis. Gene Ontology-Biological Process (GO-BP) enrichment analysis was performed to explore the potential biological functions of FXN in COAD. Functional assays (CCK-8, colony formation, Transwell) were used to confirm the critical role of FXN in COAD cell proliferation, migration, and invasion. R language, single‑sample gene-based enrichment analysis (ssGSEA), and correlation analysis were applied to identify associations between FXN expression and immune infiltration, immunotherapy response, and drug sensitivity.
Results
Compared with normal colon tissue, FXN expression was upregulated in COAD tissue (p < 0.05), and high FXN expression predicted poor prognosis in COAD patients (p < 0.05). GO-BP enrichment analysis revealed that FXN correlated with cell proliferation and microtubule dynamics (p < 0.05). Downregulating FXN inhibited proliferation, migration, and invasion of COAD cells (p < 0.05). FXN expression showed a correlation with several immune cells and immune cell markers, but no significant correlation with immunophenoscore (IPS). Drug sensitivity analysis indicated that FXN was associated with COAD samples’ sensitivity to several anticancer drugs (p < 0.001).
Conclusions
FXN may promote the malignant phenotype of COAD and correlate with immune infiltration, serving as a prognostic biomarker and a potential reference for chemotherapy but not for immune checkpoint blockade.
Keywords: Colon adenocarcinoma, FXN, Prognosis, Immune infiltration, Drug sensitivity
Introduction
Colorectal cancer (CRC) is one of the most common malignant tumors worldwide and in China, and has high incidence and mortality [1]. According to the national cancer report released by the National Cancer Center in 2024, the incidence rate and mortality rate of CRC in China in 2022 ranked second and fifth, respectively, in malignant tumors [2]. Moreover, the incidence of CRC among young people is also gradually increasing [3]. Based on histologic features, colon cancer is subdivided into colon adenocarcinoma (COAD), squamous carcinoma, adenosquamous carcinoma, undifferentiated carcinoma, etc. More than 95% of colon cancers are COAD, which originates from the mucus-secreting glands in the intestinal wall [4].
CRC is often asymptomatic. By the time symptoms such as rectal bleeding, anemia, or abdominal pain appear, most patients are already in advanced stages, at which point the cancer exhibits high invasiveness, malignancy, and metastatic potential [5]. The standard treatment for early-stage CRC involves surgery with or without adjuvant chemotherapy [6]. Chemotherapy, targeted therapy, and immunotherapy are the primary systemic treatments for patients with advanced metastatic CRC [7]. Despite continuous improvements in CRC screening and treatment methods [8, 9], the long-term prognosis for patients with CRC remains poor, and developing rational and effective treatments remains a significant challenge [10].
Frataxin (FXN) is a commonly expressed mitochondrial protein. It has been shown that FXN is involved in the biogenesis of iron-sulfur clusters, regulates the activity of iron-sulfur proteins, and participates in various cellular processes such as mitochondrial oxidative respiration [11]. A trinucleotide repeat mutation in the FXN gene causes Friedreich’s ataxia, a disorder characterized by cardiomyopathy and neurodegeneration [12]. FXN deficiency leads to mitochondrial iron accumulation, increased ROS production and lipid peroxidation [13, 14]. It has also been shown that FXN is involved in hypoxia-induced tumor stress [15]. And cells with deficient FXN expression are more prone to stress-induced mitochondrial damage and apoptosis, while the opposite is true for cells overexpressing FXN [16]. The most recent information on FXN is the role of variants in cancer. Indeed, when examining the cancer somatic mutation catalog (COSMIC database [17]), missense variants were found in several human cancer tissues [18]. Studies in related tumors have shown that elevated FXN in ovarian cancer stem cells inhibits ROS accumulation, promotes ferroptosis resistance, and regulates the antioxidant protein PRDX3 [19]. Previous studies have demonstrated that FXN can inhibit the proliferation of CRC cells [20]. However, to date, the expression of FXN in COAD and its effects on migration and invasion remain unclear.
Chemotherapy and immunotherapy play a crucial role in the treatment of COAD [21, 22]. In particular, chemotherapy drugs can be employed both as direct treatment for patients with COAD and as neoadjuvant therapy to enhance the surgical cure rate for this cancer [23, 24]. Immunotherapy is considered a promising strategy for treating or even curing certain types of cancer [25, 26]. However, the effects of FXN expression in COAD on sensitivity to chemotherapy drugs and on immune cell infiltration remain unknown at present.
This study has the goal of looking into the expression of FXN in COAD, its role in the carcinogenesis and development of COAD, and its impact on immunotherapy and chemotherapy.
Materials and methods
Patients and tissue samples
All of the clinical tissue specimens used in this study were obtained from the Second Affiliated Hospital of Zhengzhou University and verified by histopathology. Immunohistochemical staining was performed on 80 samples of COAD tissue and their paired adjacent normal colonic tissue. The study strictly adhered to all relevant ethical guidelines for human subjects.
Cell lines and culture
The human COAD cell lines RKO and SW480, and the normal colonic epithelial cell line NCM460, were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). RKO, SW480, and NCM460 cells were cultured in 1640 medium. All media were supplemented with 10% fetal bovine serum (FBS) (Genial Biologicals, Inc.). The cells were cultured in a humidified environment at 37 °C with 5% CO₂.
siRNA transfection
The COAD cells were transfected transiently with siRNA (siNC and siFXN). Twenty-four hours prior to transfection, the cells to be transfected were seeded at a density of approximately 60–80 in six-well plates. The siRNA was diluted in jetPRIME buffer according to the manufacturer’s instructions and then mixed gently. The jetPRIME® reagent was added (Polyplus Transfection®, France) and incubated at room temperature for 10 min. Finally, the transfection mixture was added to the cells in a medium containing serum, and incubated for 24–72 h. The first FXN-targeting siRNA used in this study had the following sequences: sense 5'-GAACCUAUGUGAUCAACAAGC-3' and antisense 5'-UUGUUGAUCACAUAGGUUCCU-3'. The second FXN-targeting siRNA used in this study had the following sequences: sense 5'-CCUAUACCUGAAUAUAACAAC-3' and antisense 5'-UGUUAUAUUCAGGUAUAGGCA-3'. The negative control siRNA sequences were sense 5'-UUCUCCGAACGUGUCACGUTT-3' and antisense 5'-ACGUGACACGUUCGGAGAATT-3'.
Cell counting kit‑8 assay
The COAD cells were resuspended after digestion and subjected to cell counting. The cells were then seeded into a 96-well plate at a density of 2 × 103 cells per well. Absorbance values at 450 nm were measured using the CCK-8 assay at 24, 48, 72 and 96 h over four consecutive days.
Western blotting
After lysing the cells using a radio-immunoprecipitation assay buffer (Cwbio, Jiangsu, China), protein concentrations were determined with the bicinchoninic acid protein assay kit (Beyotime, Shanghai, China). Protein samples (20 μg/lane) were separated on a 15% gel by sodium lauryl sulfate–polyacrylamide gel electrophoresis and transferred to a PVDF membrane. The membrane was blocked with 10% skim milk at room temperature for 2 h, followed by incubation overnight at 4 °C with primary antibodies against FXN (1:300; 14,147–1-AP; Proteintech, China) and GAPDH (1:10,000, P04406, CUSABIO, China) overnight at 4 °C.
Quantitative real-time PCR (qRT‒PCR)
Cell RNA was extracted using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s instructions and cDNA was synthesized using a reverse transcription kit (Vazyme, Nanjing). The qRT-PCR reaction system had a total volume of 10 µL and employed a SYBR Green qPCR master mix (Vazyme, Nanjing). The reaction was detected using the Applied ABI QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific, Inc.). The primer pairs used for qRT-PCR were: FXN forward primer 5'-GGAAACGCTGGACTCTTTAGC-3' and reverse primer 5'-CCAGTTTGACAGTTAAGACACCA-3'; GAPDH forward primer 5'-GGAGCGAGATCCCTCCAAAAT-3' and reverse primer 5'-GGCTGTTGTCATACTTCTCATGG-3'. The relative expression was normalized using the 2-ΔΔCt method and the internal reference gene GAPDH.
Colony formation assay
A colony formation assay was performed to evaluate the clonogenic capacity of COAD cells following FXN knockdown. Transfected siNC and siFXN cells were seeded at a density of approximately 1000 cells per well in six-well plates. Fifteen days later, the cells were fixed with a 4% polyformaldehyde solution and stained with a 0.1% crystal violet solution (Solis BioDyne, Beijing, China). The colonies were then washed with distilled water, photographed and counted using ImageJ software.
Transwell migration and invasion assays
Cell migration and invasion assays were performed using 24-well Transwell chambers with a pore size of 8 μm (Biofil, China). 600 µL of complete culture solution was added to the 24-well plate and then the prepared chambers were placed into the plate. Matrix gel (100 µL) was added to the upper chamber for invasion assays but not for migration assays. A suspension of transfected cells was prepared in serum-free medium. 200 µL of the cell suspension was added to each upper chamber, ensuring that each chamber contained 4 × 103 cells. The plate was placed in an incubator and incubated for 24 h, after which the cells were fixed with 4% polyformaldehyde and stained with 1% crystal violet. After staining, the cells in the upper layer of the chambers were wiped clean with a cotton swab. The cells in the lower layer of the chambers were the invading or migrating cells. The cells were photographed under an inverted microscope and counted.
Immunohistochemistry (IHC) assay
The 80 COAD specimens were sectioned into 4-micron slices and mounted on slides for microscopic observation. The tissue sections on the slides were then subjected to dewaxing and rehydration using xylene and a gradient concentration of ethanol. Next, the slides were immersed in boiling Tris/ethylenediaminetetraacetic acid (pH 9.0) solution for 20 min for antigen retrieval. Subsequently, the slides were immersed in 3% hydrogen peroxide solution for 10 min to inhibit endogenous peroxidase. Then, the slides were incubated with a primary antibody against FXN (1:300; 14,147–1-AP; Proteintech, China), followed by the HRP secondary antibody (1:1000, RGAR011; Proteintech, China), and washed three times with phosphate-buffered saline. Finally, the sections were stained with 3,3'-diaminobenzidine and substrate chromogen (Dako) for 2 min at room temperature and then counterstained with hematoxylin for 40 s. Slides incubated only with the secondary antibody without the primary antibody were used as the negative control. The staining of each section was evaluated by two independent researchers blinded to the clinicopathological parameters. Immunoreactivity for FXN expression was evaluated using the semi-quantitative immunoreactivity score (IRS) system, according to the intensity of staining (IS) (0-negative; 1-weak; 2-moderate; 3-strong) and area of positivity (AP) (0, 0%; 1, ≤ 10%; 2, 11–50%; 3, 51–80%; 4, ≥ 80%;). The final IRS was obtained by multiplying the IS and AP. In this study, FXN expression was regarded as low expression when the score was < 2, and high expression when the score was ≥ 2. The immunohistochemical staining was assessed by two separate pathologists who were blinded to patients’ information.
TCGA and GEO data collection and processing
The RNA-sequencing data of COAD and the related clinical data were obtained from The Cancer Genome Atlas (TCGA) (https://portal.gdc.cancer.gov/repository) [27]. TCGA–COAD was used for the investigation of FXN expression and its prognostic value, Gene Ontology–Biological Process (GO-BP) enrichment analysis, immune-related analysis, and drug sensitivity. Microarray data of GSE39582, GSE41258, GSE37364, GSE44076, and GSE773360 were retrieved from Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/) [28–32]. GSE39582 was applied for FXN expression and GO-BP enrichment analysis. GSE39582, GSE41258, GSE37364 and GSE44076 were analyzed for FXN expression. GSE17538 was taken for survival analysis.
GO-BP enrichment analyses
Spearman’s test was conducted using R software based on TCGA-COAD and GSE39582. FXN-related genes (p < 0.05 and the absolute value of the correlation coefficient ≥ 0.30) were subjected to GO-BP enrichment analyses using the Metascape database (http://metascape.org/).
Immune infiltration analysis
Single-sample gene set enrichment analysis (ssGSEA) was applied for the determination of immune infiltration based upon the immune gene sets [33].
Evaluation of drug sensitivity
The R language and the “oncopredict” package were employed to calculate the sensitivity score of each sample in the TCGA-COAD dataset for 199 anticancer drugs, with a higher sensitivity score indicating lower drug sensitivity.
Calculation of the scores for immune checkpoint inhibitor therapy
The immunophenoscore (IPS), a well-validated biomarker that predicts response to immune checkpoint inhibitor (ICI) therapy, was obtained from The Cancer Immunome Atlas (TCIA) (https://tcia.at/patients). For each sample in the TCGA-COAD dataset, four types of IPS were calculated, i.e., IPS-CTLA4-neg-PD1-neg, IPS-CTLA4-neg-PD1-pos, IPS-CTLA4-pos-PD1-neg, and IPS-CTLA4-pos-PD1-pos. These scores predict the potential efficacy of different treatment scenarios, i.e., the absence of both CTLA-4 and PD-1 blockade, PD-1 blockade alone, CTLA-4 blockade alone, and combined CTLA-4 and PD-1 blockade, respectively.
Statistical analysis
The statistical analysis was performed using SPSS 25.0 (SPSS Inc., USA) and GraphPad Prism 8.0 (GraphPad Software Inc., USA). All data are shown as the mean ± standard deviation (SD). Two groups of data were analyzed by independent-sample t-test or paired-sample t-test, and three or more groups of data were analyzed by one-way analysis of variance. If the data do not meet the normal distribution or the variance is not uniform, then non-parametric tests are used. Kaplan–Meier and log-rank were used to analyze the effect of FXN on the prognosis of COAD. Cox regression analysis was performed for univariate and multivariate analyses of FXN and clinicopathological parameters. Variables with a p value less than 0.1 in univariate regression were included in the multivariate regression. Pearson correlation was used for normally distributed samples, while Spearman correlation was applied for non-normally distributed ones. Categorical data are presented as frequencies and percentages (%) and inter-group comparisons are conducted using the χ2 test. p < 0.05 was considered statistically significant.
Results
FXN expression is upregulated in COAD
To investigate the expression of FXN in COAD, we first conducted relevant analyses using The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases, which showed that the mRNA expression of FXN was significantly higher in COAD tissues than in normal tissues (Fig. 1A–F, p < 0.0001). In addition, the results of western blot and qRT-PCR experiments showed that the protein and mRNA expression of FXN was higher in COAD cell lines (SW480 and RKO) than in normal colonic epithelial cell lines (NCM460; Fig. 1G and H, p < 0.05). The representative immunohistochemical results of tumor tissue and its corresponding adjacent normal tissue were shown in Fig. 1I. The final IRS analysis revealed that FXN protein was highly expressed in 68 out of 80 COAD samples (85%), compared to 37 out of 80 adjacent normal samples (46.25%). The difference is statistically significant (χ2 = 26.625, p < 0.0001). We further assessed FXN protein expression levels between COAD clinical specimens and their adjacent normal counterparts using the paired t-test, and the result was consistent with previous findings, i.e., FXN expression was significantly higher in COAD than in adjacent normal tissue (Fig. 1J, p < 0.0001).
Fig. 1.
FXN expression is upregulated in COAD. A–G Analyses based on TCGA-COAD (A), GSE39582 (B), GSE41258 (C), GSE37364 (D), GSE73360 (E), GSE44076 (F) datasets demonstrated significantly higher FXN mRNA expression in COAD cancer tissues than in paracancerous tissues. G, H Increased protein (G) and mRNA (H) levels of FXN in the COAD cell lines SW480 and RKO compared with normal colon cell line NCM460. I Representative images of IHC detection for FXN expression in cancerous and paired noncancerous samples obtained from COAD patients. The left image showed low FXN expression in normal colon tissue, while the right image demonstrated high FXN expression in COAD tissue. J Analysis of COAD clinical specimens revealed significantly higher FXN protein expression in COAD cancer tissues compared to adjacent non-cancerous tissues. All micrographs are magnified at × 200. ****p < 0.0001, ***p < 0.001, *p < 0.05.All data represent results from at least three independent experiments. (n = 3; error bar, SD)
High expression of FXN is significantly associated with poor prognosis in patients with COAD
To explore the relationship between FXN and the prognosis of COAD patients, we performed a survival analysis using the TCGA-COAD and GSE17538 datasets, and COAD clinical specimens. The results showed that high FXN expression was significantly associated with poor prognosis of COAD patients (Fig. 2A–C, p < 0.05). To assess the independent prognostic significance of FXN expression, we performed Cox regression analysis using the TCGA-COAD dataset and COAD clinical specimens. In the TCGA-COAD dataset, univariate analysis results showed that M stage, N stage, T stage, loss of MMR protein, lymphatic invasion, cancer status, postoperative treatment, venous invasion, and FXN expression were predictors of an unfavorable prognosis. Multivariate analysis indicated that M stage, N stage, lymphatic invasion, cancer status, venous invasion, and FXN expression were independently associated with a worse outcome (Fig. 2D, p < 0.05). By using COAD clinical specimens, univariate analysis identified M stage, N stage, T stage, lymphatic invasion, differentiation, venous invasion, and FXN expression as indicators of a worse prognosis. Multivariate analysis showed that M stage, differentiation, venous invasion, and FXN expression were independently correlated with shorter OS (Fig. 2E, p < 0.05).
Fig. 2.
High FXN expression in COAD tissues predicts poor prognosis. A–C Kaplan–Meier and log-rank analyses of the correlation between FXN expression and prognosis in COAD patients. D, E Univariate and multivariate Cox regression analyses based on the TCGA-COAD dataset (D) and COAD clinical specimens (E). (MMR: mismatch repair proteins)
GO-BP enrichment analysis for biological functions of FXN
To explore the specific function of FXN in the development of CRC (COAD), we performed Gene Ontology-Biological process (GO-BP) enrichment analysis using the TCGA-COAD and GSE39582 datasets. Initially, the correlation between FXN expression and other genes in the two datasets was analyzed. Genes with |rs|≥ 0.30 and p < 0.05 were selected for enrichment analysis (Fig. 3A and B). Results indicated that FXN expression was significantly correlated with cell proliferation and microtubule dynamics in COAD (Fig. 3C and D, p < 0.05).
Fig. 3.
GO-BP enrichment analysis for biological functions of FXN. A, B Correlation analysis between individual genes and FXN expression was performed using the TCGA-COAD (A) and GSE39582 (B) datasets. C, D GO-BP enrichment analysis for FXN was conducted using the TCGA-COAD dataset (C) and GSE39582 dataset (D). (rs: correlation coefficient)
FXN knockdown inhibits COAD cell malignant phenotypes
To elucidate the role of FXN in COAD, further investigation was conducted through a series of experiments. First, we knocked down the expression of FXN in the SW480 and RKO cell lines using siRNA. The efficiency of FXN knockdown was confirmed by qRT-PCR and western blot analysis (Fig. 4A and B , p< 0.05). The results of the CCK-8 cell proliferation assay and the colony formation assay showed that FXN knockdown significantly reduced cell proliferation ability (Fig. 4C, D , p < 0.05). As shown in Fig. 4E, FXN knockdown significantly attenuated the migration and invasion abilities of COAD cells (p < 0.05). In conclusion, these results indicated that in vitro, FXN knockdown significantly attenuated the proliferation, migration and invasion abilities of COAD cells.
Fig. 4.
FXN knockdown inhibits COAD cell malignant phenotypes. A, B Downregulation of FXN expression reduced its mRNA (A) and protein (B) levels in SW480 and RKO cell lines. C The proliferation of two groups of COAD cells transfected with siNC or siFXN was assessed using the CCK-8 assay. D The effect of knocking down FXN expression on COAD cell colony formation. E, F The Transwell assay was used to evaluate the effects of FXN knockdown on the migration and invasion capabilities of SW480 (E) and RKO (F) cells. (n = 3; error bar, SD;****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05)
Correlations between FXN expression and infiltrating immune cells
Using the TCGA-COAD dataset, the correlation between immune infiltration and FXN expression was investigated via ssGSEA and Spearman’s correlation analysis. Analysis results indicated that the expression of FXN was negatively correlated with CD56dim natural killer cells, central memory CD4 T cells, monocytes and natural killer cells, while it was positively correlated with activated CD4 cells, activated CD8 T cells and Type 2 T helper cells. (Table 1, p < 0.05). To validate these findings, correlations between FXN expression and immune cell marker expression were examined. The results indicated that the expression of FXN was negatively correlated with markers of monocytes and natural killer cells, and positively correlated with markers on activated CD8 T cells, which was consistent with the above assays. However, FXN expression was inversely correlated with markers of Type 2 T helper cells, which was contrary to the results of the above analysis. (Table 2, p < 0.05).
Table 1.
Correlation between FXN expression and immune infiltration in COAD
| Immune cells | Correlation | p value |
|---|---|---|
| Activated B cell | 0.025 | 0.644 |
| Activated CD4 T cell | 0.322 | < 0.001 |
| Activated CD8 T cell | 0.187 | 0.001 |
| Activated dendritic cell | 0.018 | 0.749 |
| CD56bright natural killer cell | − 0.063 | 0.253 |
| CD56dim natural killer cell | − 0.117 | 0.032 |
| Central memory CD4 T cell | − 0.114 | 0.037 |
| Central memory CD8 T cell | − 0.051 | 0.353 |
| Effector memory CD4 T cell | − 0.054 | 0.327 |
| Effector memory CD8 T cell | 0.007 | 0.899 |
| Eosinophil | − 0.083 | 0.128 |
| Gamma delta T cell | 0.008 | 0.887 |
| Immature B cell | − 0.011 | 0.837 |
| Immature dendritic cell | − 0.013 | 0.818 |
| Macrophage | − 0.041 | 0.451 |
| Mast cell | − 0.085 | 0.118 |
| MDSC | − 0.025 | 0.643 |
| Memory B cell | − 0.092 | 0.091 |
| Monocyte | − 0.135 | 0.013 |
| Natural killer cell | − 0.153 | 0.005 |
| Natural killer T cell | − 0.071 | 0.191 |
| Neutrophil | − 0.009 | 0.867 |
| Plasmacytoid dendritic cell | − 0.091 | 0.094 |
| Regulatory T cell | − 0.09 | 0.101 |
| T follicular helper cell | − 0.029 | 0.595 |
| Type 1 T helper cell | − 0.025 | 0.653 |
| Type 17 T helper cell | 0.104 | 0.056 |
| Type 2 T helper cell | 0.199 | < 0.001 |
Table 2.
Correlation between FXN expression and immune cell markers in COAD
| Immune cells | Gene | Correlation | p value |
|---|---|---|---|
| CD8 + T cell | CD8A | 0.189 | < 0.001 |
| CD8B | 0.141 | 0.009 | |
| PTPRC | 0.017 | 0.759 | |
| T cell (general) | CD3D | 0.1 | 0.067 |
| CD3E | 0.06 | 0.273 | |
| CD2 | 0.154 | 0.005 | |
| B cell | CD19 | 0.023 | 0.677 |
| CD79A | − 0.012 | 0.82 | |
| CD27 | 0.003 | 0.951 | |
| KRT20 | − 0.031 | 0.576 | |
| Monocyte | CCL2 | − 0.023 | 0.671 |
| CD14 | − 0.061 | 0.268 | |
| CD68 | − 0.153 | 0.005 | |
| IL10 | 0.027 | 0.62 | |
| CSF1R | − 0.149 | 0.006 | |
| CD86 | 0.021 | 0.7 | |
| TAM | CCL2 | − 0.023 | 0.671 |
| CD68 | − 0.153 | 0.005 | |
| IL10 | 0.027 | 0.62 | |
| M1 Macrophage | NOS2 | 0.15 | 0.006 |
| CD80 | 0.056 | 0.31 | |
| IRF5 | − 0.067 | 0.218 | |
| IL6 | − 0.041 | 0.454 | |
| FCGR1A | 0.009 | 0.869 | |
| PTGS2 | − 0.059 | 0.278 | |
| M2 Macrophage | CD163 | − 0.046 | 0.398 |
| MRC1 | − 0.04 | 0.461 | |
| VSIG4 | − 0.01 | 0.851 | |
| MS4A4A | 0.04 | 0.463 | |
| Neutrophil | CEACAM8 | − 0.028 | 0.604 |
| ITGAM | − 0.158 | 0.004 | |
| FUT4 | 0.279 | < 0.001 | |
| CCR7 | − 0.091 | 0.095 | |
| Natural killer cell | KIR2DS4 | 0.099 | 0.069 |
| NCAM1 | − 0.283 | < 0.001 | |
| NCR1 | 0.027 | 0.621 | |
| Dendritic cell | CD1C | − 0.12 | 0.028 |
| NRP1 | − 0.094 | 0.087 | |
| THBD | − 0.234 | < 0.001 | |
| IL3RA | − 0.216 | < 0.001 | |
| ITGAX | − 0.122 | 0.026 | |
| Th1 | TBX21 | 0.089 | 0.105 |
| IFNG | 0.183 | 0.001 | |
| STAT4 | 0.08 | 0.143 | |
| STAT1 | 0.133 | 0.015 | |
| Th2 | GATA3 | − 0.08 | 0.143 |
| IL17A | 0.045 | 0.411 | |
| STAT6 | − 0.142 | 0.009 | |
| Tfh | BCL6 | − 0.199 | < 0.001 |
| Th17 | STAT3 | 0.036 | 0.514 |
| IL17A | 0.045 | 0.411 | |
| Treg | FOXP3 | − 0.079 | 0.149 |
| IL2RA | 0.042 | 0.446 | |
| CCR8 | − 0.097 | 0.077 | |
| STAT5B | − 0.201 | < 0.001 | |
| TGFB1 | − 0.083 | 0.129 | |
| T cell exhaustion | PDCD1 | 0.098 | 0.073 |
| CTLA4 | 0.022 | 0.69 | |
| LAG3 | 0.157 | 0.004 | |
| HAVCR2 | 0.043 | 0.432 | |
| GZMB | 0.14 | 0.01 |
Impact of FXN expression on the efficacy of immune checkpoint inhibitor therapy
To elucidate the impact of FXN expression on the efficacy of ICI in COAD, this study analyzed the correlation between FXN expression and IPS using the TCIA database. Based on patients’ MSI/MMR stratification, we conducted stratified analyses. The results showed that FXN expression was not significantly correlated with any of the four IPS indicators (Table 3, p > 0.05).
Table 3.
Correlation between FXN expression and response to IPS
| MSI/MMR | IPS Scoring Metrics | Correlation | p value |
|---|---|---|---|
| MSI-H | ips_ctla4_neg_pd1_neg | − 0.154 | 0.302 |
| ips_ctla4_neg_pd1_pos | − 0.214 | 0.149 | |
| ips_ctla4_pos_pd1_neg | − 0.087 | 0.559 | |
| ips_ctla4_pos_pd1_pos | − 0.267 | 0.069 | |
| MSI-L | ips_ctla4_neg_pd1_neg | 0.130 | 0.315 |
| ips_ctla4_neg_pd1_pos | 0.083 | 0.523 | |
| ips_ctla4_pos_pd1_neg | 0.077 | 0.552 | |
| ips_ctla4_pos_pd1_pos | 0.069 | 0.596 | |
| MSS | ips_ctla4_neg_pd1_neg | 0.014 | 0.845 |
| ips_ctla4_neg_pd1_pos | 0.021 | 0.762 | |
| ips_ctla4_pos_pd1_neg | − 0.027 | 0.702 | |
| ips_ctla4_pos_pd1_pos | − 0.043 | 0.536 |
Drug sensitivity analysis
To clarify the impact of FXN expression on the efficacy of various anticancer drugs in the treatment of COAD, we analyzed the sensitivity scores of individual COAD samples in the TCGA-COAD dataset to various anticancer drugs, where higher scores indicated lower drug sensitivity, that is, reduced responsiveness. Subsequently, we explored the association between drug sensitivity and FXN expression. Results showed that FXN expression was negatively correlated with sensitivity scores for oxaliplatin, 5-fluorouracil, and irinotecan (Fig. 5A–C , p < 0.001).
Fig. 5.
Correlation between FXN expression and drug sensitivity based on the TCGA-COAD dataset. A–C Correlation between sensitivity scores for oxaliplatin (A), irinotecan (B), and 5-fluorouracil (C) and FXN mRNA expression. (rs: correlation coefficient)
Discussion
FXN is a widely expressed mitochondrial protein that plays a crucial role in Fe-S cluster synthesis. Fe-S clusters serve as protein cofactors mediating redox reactions within the electron transport chain and other pathways [34]. The mitochondrial Fe-S cluster assembly complex comprises several components: cystathionine desulfurase NFS1 and its cofactor ISD11, the assembly scaffold ISCU2, and frataxin itself [35, 36]. The interaction between FXN and the Fe-S cluster assembly complex not only enhances the efficiency of sulfur transfer but also supports the formation of stable Fe-S clusters, which are essential for mitochondrial function [34]. Previous studies demonstrated that FXN deficiency generates a senescent endothelial subpopulation, promoting vascular inflammatory and proliferative signals in other cells to drive disease [37, 38]. Research on related tumors indicated that elevated FXN in ovarian cancer stem cells suppressed ROS accumulation, promoted iron-mediated resistance, and regulated the antioxidant protein PRDX3 [18]. However, the expression and role of FXN in COAD remain unclear. This study investigated the function of FXN in COAD development and its association with immunotherapy and chemotherapy.
This study analyzed the TCGA and GEO databases and found that FXN mRNA expression was significantly elevated in COAD tissues compared to normal samples. The same results were obtained using qRT-PCR, Western blot, and IHC experiments in colon cancer cells, normal colon cells, and clinical COAD specimens. Furthermore, Kaplan–Meier analysis and the Log-Rank test revealed that high FXN expression correlated with poorer prognosis in COAD patients. Cox regression analysis indicated that elevated FXN expression was an independent predictor of adverse prognosis. Collectively, these findings suggested that FXN may play a role in promoting the development of COAD and may represent a potential prognostic biomarker for patients with this disease.
To elucidate the specific function of FXN in COAD, researchers employed GO-BP enrichment analysis to identify biological processes associated with FXN. Results indicated that FXN correlated with cell proliferation and microtubule function within COAD. Microtubules serve as critical transport tracks for organelles and molecules via kinesin-dependent movement, facilitating cell migration by delivering essential proteins and signaling molecules to the cell front. During cell migration, microtubules and actin fibers synergistically transmit mechanical forces: microtubules provide bending stiffness while actin filaments generate contractile forces, jointly ensuring efficient cell movement. Thus, we hypothesized that FXN may be involved in the regulation of proliferation and migration in COAD cells. To verify this hypothesis, researchers examined the impact of FXN on COAD cell biological behaviors.
Functional assay results indicated that FXN knockdown inhibited COAD cell proliferation, consistent with enrichment analysis findings. This study further revealed that suppressing FXN expression also inhibited biological behaviors such as migration and invasion in COAD cells, aligning with enrichment analysis outcomes. In summary, our findings indicated that elevated FXN expression correlated with malignant progression in COAD. Combined with the prognostic significance of increased FXN expression in COAD patients, the role of FXN in COAD tumorigenesis and progression has been partially elucidated. Although previous studies have demonstrated that FXN can inhibit tumor cell proliferation in the colon cancer cell line (MIP101) [15], this finding differs from our results. This discrepancy may stem from several possible causes: 1) Cell line variability: the cell lines used differ between studies. We utilized RKO and SW480 cells, whereas the previous study used MIP101 cells, which may exhibit distinct biological responses. 2) Transfection methods: different transfection strategies were employed. We used siRNA-mediated knockdown, while the previous study adopted retroviral transfection. 3) Direction of genetic manipulation: the studies manipulated FXN expression in opposite directions; we performed FXN knockdown, whereas the previous study investigated FXN overexpression. 4) Assay methodologies: the specific methods for validating phenotypes varied. We assessed proliferation, migration, and invasion using CCK-8, colony formation, and Transwell assays. In contrast, the previous study relied on population doubling time calculations and soft agar assays. Meanwhile, multiple previous studies have demonstrated that FXN possesses pro-cancer properties. For instance, in ovarian cancer patients, high FXN expression is associated with poorer progression-free survival and FXN knockdown suppresses tumor-initiating capacity [19]. High levels of NFS1 and FXN are strongly associated with poor prognosis in patients with diffuse large B-cell lymphoma. Following FXN inhibition, DLBCL cells also exhibit suppressed cell viability and increased cytotoxicity [39].
The present study also investigated the effect of FXN expression on immunotherapy in COAD. Results from ssGSEA immune infiltration analysis and correlation analysis revealed that FXN expression positively correlated with activated CD4 T cell, activated CD8 T cell and type 2 T helper cell, but was negatively associated with CD56dim natural killer cell, central memory CD4 T cell, monocyte and natural killer cell. Analysis of the correlation between FXN expression and immune cell markers revealed that FXN was negatively correlated with markers of monocytes and natural killer cells, and positively correlated with markers of activated CD8 T cells, consistent with the ssGSEA results. However, FXN expression was negatively correlated with markers of type 2 helper T cells, contrary to the results of the preceding analysis. The reason for this discrepancy may lie in the fundamental differences in their respective methodologies. ssGSEA evaluates the enrichment level of a specific immune cell type within a single sample by utilizing a predefined gene set, with results more reflective of the relative abundance and overall activity level of that cell type in the tumor microenvironment [40, 41]. The gene sets for each immune cell are relatively comprehensive. For example, the Th2 gene set comprises 28 genes. In contrast, marker-based analysis typically examines the expression of only a few widely recognized specific marker genes, such as assessing Th2 presence by measuring the expression of GATA3, IL17A, and STAT6. However, the expression of these markers can be influenced by various microenvironmental factors and is not exclusive to Th2 cells [42]. Therefore, inconsistencies between ssGSEA results and marker-based analysis are to be expected. As these results suggested that FXN might be associated with alterations in the immune microenvironment, we hypothesized that FXN could be related to the response to immunotherapy in COAD. Therefore, we subsequently analyzed the correlation between FXN expression and IPS scores. The results showed that FXN expression was not significantly correlated with IPS in any of the groups—MSI-H, MSI-L, or MSS—indicating that FXN may not serve as a predictive biomarker for the efficacy of immune checkpoint inhibitor therapy. Nevertheless, as these results were derived exclusively from bioinformatic analyses, they should be considered preliminary and require confirmation through clinical studies.
Drug sensitivity analysis indicated that COAD samples with high FXN expression exhibit high sensitivity to oxaliplatin, 5-fluorouracil, and irinotecan—all three of which are well-established chemotherapy drugs for treating COAD in clinical practice. Consequently, elevated FXN expression may enhance the efficacy of conventional clinical chemotherapy drugs against COAD. These findings provide important insights for the clinical management of COAD.
Additionally, our study has several limitations. First, the signaling pathways through which FXN influences COAD cell biology remain unclear and will be determined in future research. Second, the relationship between FXN expression and immune infiltration/immunotherapy response scores was determined using bioinformatics databases, thus, in vivo and in vitro experiments are needed to investigate FXN’s impact on immunotherapy and its underlying mechanisms. And since FXN’s influence on drug sensitivity was identified solely through bioinformatics analysis, in vivo and in vitro experiments are required to further validate the effect of FXN expression on chemotherapy drug sensitivity.
Conclusion
In summary, our research indicated that the high expression of FXN in COAD may influence tumor cell proliferation, migration, and invasive capabilities, ultimately affecting patient prognosis. Furthermore, our findings suggested that the expression level of FXN may have a certain impact on the sensitivity of commonly used chemotherapy drugs in COAD, thus providing some reference value for the clinical management of this disease. In addition, while FXN expression correlates significantly with specific immune cell infiltration, it lacks predictive value for the efficacy of immune checkpoint blockade.
Acknowledgements
The authors thank the contributors of TCGA (https://portal.gdc.cancer.gov/repository) and GEO (https://www.ncbi.nlm.nih.gov/geo/) for sharing the related data.
Abbreviations
- CCK-8
Cell counting kit-8
- COAD
Colon adenocarcinoma
- CRC
Colorectal cancer
- CTLA-4
Cytotoxic T-lymphocyte-associated protein 4
- FBS
Fetal bovine serum
- FXN
Frataxin
- GEO
Gene Expression Omnibus
- GO-BP
Gene ontology—biological process
- IHC
Immunohistochemistry
- ICI
Immune checkpoint inhibitors
- IPS
Immunophenoscore
- IRS
Immunoreactivity score
- MMR
Mismatch repair
- PD-1
Programmed cell death protein 1
- qRT-PCR
Quantitative real-time PCR
- ROS
Reactive oxygen species
- ssGSEA
Single-sample gene set enrichment analysis
- TCGA
The Cancer Genome Atlas
Author contributions
Z.W.C. and H.N. contributed to the conception and design of the study. S.M.Y. wrote the first draft of the manuscript. Z.W.C., H.N. and S.M.Y. analyzed the bioinformatical data. S.Q.Z., M.Y.H. and H.N. provided COAD samples and detected FXN expression. S.M.Y. conducted cell-based experiments and performed statistical analysis. Z.W.C. and H.N. supervised all the work. All authors contributed to the article and approved the submitted version.
Funding
This work was supported by grants from the Henan Provincial Natural Science Foundation (No. 252300421368), the Top Talents Funding of the Second Affiliated Hospital of Zhengzhou University (No. 2020BJRCB04), and Beijing Science and Technology Innovation Medical Development Foundation (No. KC2023-JX-0186-FM054), Mengya Sun.
Data availability
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.
Declarations
Conflict of interest
The authors declare that there are no competing interests.
Ethical approval
Our study has been approved by the Institutional Research Ethics Committee of the Second Affiliated Hospital of Zhengzhou University. Ethics Approval Number: KY2025433.
Informed consent
Written informed consent was provided by every patient.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Wenchao Zhao, Email: wenchao_zhao@zzu.edu.cn.
Na Han, Email: hannahappy@zzu.edu.cn.
References
- 1.Wang Z, Dan W, Zhang N, et al. Colorectal cancer and gut microbiota studies in China. Gut Microbes. 2023;15(1):2236364. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ferlay J, Ervik M, Lam F, et al. Global cancer observatory: cancer today. Lyon, France: International Agency for Research on Cancer. 2020, Available from: https://gco.iarc.fr/today, accessed date: 28 Aug 2022
- 3.Campos FG. Colorectal cancer in young adults: a difficult challenge. World J Gastroenterol. 2017;23(28):5041–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Marley AR, Nan H. Epidemiology of colorectal cancer. Int J Mol Epidemiol Genet. 2016;7(3):105–14. [PMC free article] [PubMed] [Google Scholar]
- 5.Xi Y, Xu P. Global colorectal cancer burden in 2020 and projections to 2040. Transl Oncol. 2021;14(10):101174. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Audisio A, Fazio R, Daprà V, et al. Neoadjuvant chemotherapy for early-stage colon cancer. Cancer Treat Rev. 2024;123:102676. [DOI] [PubMed] [Google Scholar]
- 7.Leowattana W, Leowattana P, Leowattana T. Systemic treatment for metastatic colorectal cancer. World J Gastroenterol. 2023;29(10):1569–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schreuders EH, Ruco A, Rabeneck L, et al. Colorectal cancer screening: a global overview of existing programmes. Gut. 2015;64(10):1637–49. [DOI] [PubMed] [Google Scholar]
- 9.Block KI, Block PB, Gyllenhaal C. Integrative treatment for colorectal cancer: a comprehensive approach. J Altern Complement Med. 2018;24(9–10):890–901. [DOI] [PubMed] [Google Scholar]
- 10.Shi J, Sun Z, Gao Z, et al. Radioimmunotherapy in colorectal cancer treatment: present and future. Front Immunol. 2023;14:1105180. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Culley MK, Rao RJ, Mehta M, et al. Frataxin deficiency disrupts mitochondrial respiration and pulmonary endothelial cell function. Vascul Pharmacol. 2023;151:107181. 10.1016/j.vph.2023.107181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Koutnikova H, Campuzano V, Foury F, et al. Studies of human, mouse and yeast homologues indicate a mitochondrial function for frataxin. Nat Genet. 1997;16(4):345–51. [DOI] [PubMed] [Google Scholar]
- 13.Huang ML, Becker EM, et al. Elucidation of the mechanism of mitochondrial iron loading in Friedreich’s ataxia by analysis of a mouse mutant. Proc Natl Acad Sci USA. 2009;106:16381–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Guccini I, Serio D, Condò I, et al. Frataxin participates to the hypoxia-induced response in tumors. Cell Death Dis. 2011;2(2):e123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Schulz TJ, Thierbach R, Voigt A, et al. Induction of oxidative metabolism by mitochondrial frataxin inhibits cancer growth: Otto Warburg revisited. J Biol Chem. 2006;281(2):977–81. [DOI] [PubMed] [Google Scholar]
- 16.Condò I, Ventura N, Malisan F, et al. A pool of extramitochondrial frataxin that promotes cell survival. J Biol Chem. 2006;281(24):16750–6. [DOI] [PubMed] [Google Scholar]
- 17.Wang Y, Zhao G, Condello S, et al. Frizzled-7 identifies platinum-tolerant ovarian cancer cells susceptible to ferroptosis. Cancer Res. 2021;81:384–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Huang Y, Lin J, Xiong Y, et al. Superparamagnetic iron oxide nanoparticles induce ferroptosis of human ovarian cancer stem cells by weakening cellular autophagy. J Biomed Nanotechnol. 2020;16:1612–22. [DOI] [PubMed] [Google Scholar]
- 19.Xu S, Liu Y, Yang S, et al. FXN targeting induces cell death in ovarian cancer stem-like cells through PRDX3-mediated oxidative stress. iScience. 2024;27(8):110506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Sharma P, Wagner K, Wolchok JD, et al. Novel cancer immunotherapy agents with survival benefit: recent successes and next steps. Nat Rev Cancer. 2011;11:805–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Modest DP, Pant S, Sartore-Bianchi A. Treatment sequencing in metastatic colorectal cancer. Eur J Cancer. 2019;1 09:70–83. [DOI] [PubMed] [Google Scholar]
- 22.Abedizadeh R, Majidi F, Khorasani HR, et al. Colorectal cancer: a comprehensive review of carcinogenesis, diagnosis, and novel strategies for classified treatments. Cancer Metastasis Rev. 2024;43(2):729–53. [DOI] [PubMed] [Google Scholar]
- 23.Smith HG, Nilsson PJ, Shogan BD, et al. Neoadjuvant treatment of colorectal cancer: comprehensive review. BJS Open. 2024;8(3):zrae038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Fakih MG. Metastatic colorectal cancer: current state and future directions. J Clin Oncol. 2015;33(16):1809–24. [DOI] [PubMed] [Google Scholar]
- 25.Riley RS, June CH, Langer R, et al. Delivery technologies for cancer immunotherapy. Nat Rev Drug Discov. 2019;18(3):175–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Salva de Torres C, Baraibar I, Saoudi González N, et al. Current and emerging treatment paradigms in colorectal cancer: integrating hallmarks of cancer. Int J Mol Sci. 2024;25(13):6967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Weinstein JN, Collisson EA, Mills GB, et al. The Cancer Genome Atlas Pan-Cancer analysis project. Nat Genet. 2013;45(10):1113–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Marisa L, de Reyniès A, Duval A, et al. Gene expression classification of colon cancer into molecular subtypes: characterization, validation, and prognostic value. PLoS Med. 2013;10(5):e1001453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sheffer M, Bacolod MD, Zuk O, et al. Association of survival and disease progression with chromosomal instability: a genomic exploration of colorectal cancer. Proc Natl Acad Sci U S A. 2009;106(17):7131–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Valcz G, Patai AV, Kalmár A, Péterfia B, et al. Myofibroblast-derived SFRP1 as potential inhibitor of colorectal carcinoma field effect. PLoS ONE. 2014;9(11):e106143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Solé X, Crous-Bou M, Cordero D, et al. Discovery and validation of new potential biomarkers for early detection of colon cancer. PLoS ONE. 2014;9(9):e106748. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Smith JJ, Deane NG, Wu F, et al. Experimentally derived metastasis gene expression profile predicts recurrence and death in patients with colon cancer. Gastroenterology. 2010;138(3):958–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Charoentong P, Finotello F, Angelova M, et al. Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint blockade. Cell Rep. 2017;18(1):248–56. [DOI] [PubMed] [Google Scholar]
- 34.Dong YN, Mercado-Ayón E, Coulman J, et al. The regulation of the disease-causing gene FXN. Cells. 2024;13(12):1040. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Castro IH, Pignataro MF, Sewell KE, et al. Frataxin structure and function. Subcell Biochem. 2019;93:393–438. [DOI] [PubMed] [Google Scholar]
- 36.Zhang W, Xu L, Zhao H, et al. Mammalian mitochondrial iron-sulfur cluster biogenesis and transfer and related human diseases. Biophys Rep. 2021;7:127–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Culley MK, Zhao J, Tai YY, et al. Frataxin deficiency promotes endothelial senescence in pulmonary hypertension. J Clin Invest. 2021;131(11):e136459. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Gao J, Huang C, Kong L, et al. SIRT3 regulates clearance of apoptotic cardiomyocytes by deacetylating frataxin. Circ Res. 2023;133(7):631–47. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Shi X, Zhao Y, Gao HY, et al. NFS1, together with FXN, protects cells from ferroptosis and DNA damage in diffuse large B-cell lymphoma. Redox Biol. 2025;87:103878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Wang L, Wang L, He P. Comprehensive analysis of immune-related gene signature based on ssGSEA algorithms in the prognosis and immune landscape of hepatocellular carcinoma. Front Genet. 2022;13:1064432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Barbie DA, Tamayo P, Boehm JS, et al. Systematic RNA interference reveals that oncogenic KRAS-driven cancers require TBK1. Nature. 2009;462(7269):108–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Kaushik D, Xu B, Kumar M. Biomarkers in immunology: their impact on immune function and response. Adv Biomark Sci Tech. 2025;7:95–110. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.