RNA-Binding Protein CSTF2 Drives Immune Evasion and Malignant Progression in Breast Cancer by Degrading THSD1 mRNA

Abstract

Objective

Resistance to immunotherapy in breast cancer (BC) is intimately connected to the dynamic tumor immunoediting process. RNA-binding proteins (RBPs) are recognized as pivotal modulators of BC advancement. Cleavage stimulation factor 2 (CSTF2), a core factor involved in mRNA 3′-end maturation, shows significant overexpression in numerous cancers and is correlated with adverse clinical outcomes. However, the precise molecular mechanisms through which it drives malignancy by regulating immunoediting remain incompletely understood.

Methods

This study performed a systematic evaluation of CSTF2’s expression profile and prognostic significance in BC using data from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO). CSTF2 expression was modulated via knockdown and overexpression approaches to study core malignant traits, such as proliferative capacity and cell migration. The role of CSTF2 in immunoediting was examined through immune landscape analysis and flow cytometry.

Results

Integrated bioinformatics identified a regulatory pathway whereby CSTF2 induces transcriptional upregulation of the immune checkpoint molecule programmed death-ligand 1 (PD-L1) by promoting degradation of thrombospondin type-1 domain-containing protein 1 (THSD1) mRNA. The consequent increase in PD-L1 expression potently inhibits CD8⁺ T cell-mediated tumoricidal activity, thereby facilitating immune escape and collectively accelerating malignant progression in BC.

Conclusion

This study delineates the detailed molecular mechanisms by which the RBP CSTF2 exacerbates malignant progression, offering fresh theoretical insights and potential targets for targeted immunotherapeutic strategies in BC.

Introduction

Breast cancer (BC) remains the most prevalent malignant tumor worldwide and is a leading cause of cancer-related deaths in women.¹ Contemporary classification systems define four principal molecular subtypes based on the expression status of key biomarkers: estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and the proliferation marker Ki-67.² Among these, triple-negative breast cancer (TNBC)—defined by the absence of ER, PR, and HER2 expression—is linked with a particularly poor prognosis. This unfavorable outcome is largely due to a scarcity of effective molecular therapeutic targets and the inherently aggressive clinical behavior of this subtype, characterized by high metastasis and recurrence rates.^3,4 For patients with ERα-positive (ER+) BC, adjuvant treatment protocols still incorporate the selective estrogen receptor modulator tamoxifen as a cornerstone agent.^5,6 However, the emergence of acquired resistance during therapy represents a major clinical challenge. In this context, immune checkpoint inhibitors (ICIs) targeting the PD-1/PD-L1 axis have surfaced as a promising treatment, demonstrating significant clinical benefit for a subset of patients. Nonetheless, their overall efficacy remains limited, with objective response rates consistently below 20%.⁷ A key factor underlying this therapeutic resistance is the tumor’s inherent capacity to undergo immunoediting—an adaptive, dynamic process in which cancer cells progressively remodel their microenvironment to evade immune detection while simultaneously establishing an immunosuppressive niche that supports their survival and expansion.⁸ Therefore, identifying intrinsic cellular drivers that initiate and sustain immunoediting is crucial for developing innovative strategies to overcome this therapeutic obstacle.

RNA-binding proteins (RBPs) have surfaced as critical, yet often underexplored, regulators of various aspects of tumor biology. By engaging in selective interactions with target mRNAs, RBPs exert fine-tuned control over key post-transcriptional events, including RNA stability, translational efficiency, and subcellular localization.⁹ This layer of post-transcriptional governance significantly influences a broad spectrum of oncogenic phenotypes, encompassing dysregulated proliferation, evasion of programmed cell death, cellular senescence, enhanced motility and invasiveness, and the promotion of angiogenesis. Consequently, RBPs fulfill essential roles throughout the continuum of tumor initiation, progression, and clinical outcome.¹⁰ Cleavage stimulation factor 2 (CSTF2), a core component of the canonical mRNA 3′-end processing machinery, serves as a prime example of this important regulatory class. CSTF2 has been documented to be aberrantly overexpressed in a variety of human malignancies.¹¹ In hepatocellular carcinoma (HCC), elevated CSTF2 expression correlates with advanced histological grade and poor prognosis, a relationship mechanistically tied to the reprogramming of cellular metabolism towards increased glycolytic flux.¹² In pancreatic ductal adenocarcinoma (PDAC), CSTF2-mediated N⁶-methyladenosine (m⁶A) modifications exert positive regulatory effects on host gene expression, underscoring its potential as a target in precision oncology.¹³ Notably, pharmacological inhibition of CSTF2’s RNA recognition motif has been shown to enhance anti-tumor immune responses and surmount resistance to immune checkpoint blockade therapy, highlighting its therapeutic relevance in immunotherapy.¹⁴ In BC, CSTF2 is transcriptionally upregulated by epidermal growth factor (EGF) receptor signaling and functions as a key mediator of widespread 3′-untranslated region (3′-UTR) shortening, thereby contributing to disease pathogenesis.¹⁵ Despite these insights, its specific functional contribution to regulating the breast tumor immune microenvironment, the comprehensive network of its downstream target genes, and its precise mechanisms of action within the immunoediting framework remain to be fully elucidated. Moreover, it is still unclear whether CSTF2 promotes immunoediting specifically via the targeted degradation of particular mRNA substrates.

Thrombospondin type-1 domain-containing protein 1 (THSD1) is a secreted extracellular matrix glycoprotein engaged in fundamental biological processes such as cell-substrate adhesion and intercellular communication. THSD1 functions as a molecular bridge between the extracellular matrix and intracellular immune regulatory pathways, and genetic aberrations in THSD1 have been causally linked to developmental disorders of the vascular system.¹⁶ A growing body of evidence emphasizes the pathophysiological significance of THSD1 in oncogenesis, with documented functional roles in lung,¹⁷ colorectal,¹⁸ and esophageal cancers.¹⁹ Mechanistically, THSD1 has been reported to participate in activating transforming growth factor-beta (TGF-β) signaling pathways and has been implicated in the pathogenesis of intracranial aneurysms.²⁰ Despite increasing recognition of its functional contributions to cancer biology and vascular pathophysiology, the upstream regulatory mechanisms governing THSD1 expression—particularly the identity and mode of action of RBPs that dictate THSD1 transcript fate during breast tumor development—remain poorly defined. The present investigation is designed to address this critical knowledge gap by elucidating the specific molecular mechanisms through which an RBP regulates THSD1 expression, thereby probing a novel and potentially targetable regulatory axis in BC biology.

Materials and Methods

Protein Extraction and Western Blot

Cellular proteins were isolated with RIPA lysis buffer (Beyotime, Shanghai, China) containing a cocktail of protease and phosphatase inhibitors. Protein concentrations were determined using a standard assay, and equal amounts (30 µg per lane) were subjected to electrophoresis on 10% SDS-polyacrylamide gels. Following separation, proteins were electrotransferred onto 0.45 µm PVDF membranes (Millipore, Burlington, MA, USA). The membranes were then blocked to minimize non-specific antibody binding prior to incubation with primary antibodies. Primary antibodies included: rabbit monoclonal anti-CSTF2 (1:1000; abcam, Cambridge, UK, ab138486), rabbit polyclonal anti-THSD1 (1:1000; abcam, ab156573), rabbit monoclonal anti-Ki-67 (1:500; abcam, ab21700), rabbit monoclonal anti-Proliferating Cell Nuclear Antigen (PCNA; 1:1000; abcam, ab92552), rabbit monoclonal anti-programmed death-ligand 1 (PD-L1; 1:1000; abcam, ab205921), and mouse monoclonal anti-β-actin (1:1000; abcam, ab8226) serving as a loading control. Following washes, membranes were incubated with appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies. Immunoreactive bands were visualized using an enhanced chemiluminescence substrate (Meilunbio, Dalian, China) and detected with a chemiluminescence imaging system.

Cell Culture

The human mammary epithelial cell line MCF-10A and a panel of BC cell lines (MCF7, T47D, MDA-MB-468, MDA-MB-231, HS578T) as well as the mouse BC cell line 4T1 were acquired from the American Type Culture Collection (USA). The cell lines were verified by short tandem repeat (STR) profiling and tested negative for Mycoplasma. MCF-10A cells were maintained in DMEM/F12 medium (Gibco, USA) supplemented with 0.5 µg/mL hydrocortisone (Sigma-Aldrich, St. Louis, MO, USA), 10 µg/mL insulin (Sigma-Aldrich), 20 ng/mL recombinant human EGF (Peprotech, Cranbury, NJ, USA), 1% penicillin-streptomycin (Biosharp, Hefei, China), and 10% fetal bovine serum (FBS; Absin, Shanghai, China). The cell lines were grown in RPMI-1640 or DMEM (Gibco) containing 10% FBS, and all were incubated at 37°C in a humidified atmosphere with 5% CO₂.

Cell Transfection

Both loss-of-function and gain-of-function strategies were employed to modulate endogenous CSTF2 expression. CSTF2-specific small interfering RNAs (siRNAs; siCSTF2-1 and siCSTF2-2), a scrambled non-targeting control siRNA (siNC), a CSTF2 overexpression plasmid (pcDNA3.1-CSTF2), and its corresponding empty vector (pcDNA3.1) were commercially obtained (RiboBio Co., Ltd., Guangzhou, China) and sequence-verified. The final concentration for transfection was 50 nM. Transient transfection of T47D cells was performed using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA) following the manufacturer’s guidelines. Cells were harvested after 48 h of incubation. Knockdown or overexpression efficiency was assessed by quantitative real-time polymerase chain reaction (RT-qPCR) and Western blot.

In vivo Experiments

A subcutaneous xenograft model was utilized to evaluate the tumorigenic potential of CSTF2. All procedures involving animals received approval from the Institutional Animal Care and Use Committee (IACUC) of the Medical Ethics Committee of The Central Hospital of Yongzhou. Male BALB/c mice (6–8 weeks old, 18–20 g) were procured from Chongqing Enswell Biotechnology Co., Ltd. and maintained under specific pathogen-free conditions. Mice were randomly allocated into two groups (n = 6 per group): one inoculated with 4T1 cells expressing a control shRNA (shNC), and the other with 4T1 cells stably expressing CSTF2-targeting shRNA (shCSTF2). For inoculation, exponentially growing cells were trypsinized, washed, and resuspended in sterile phosphate-buffered saline (PBS). Each mouse received a subcutaneous injection of 1 × 10⁶ cells in 0.2 mL PBS into the left dorsal flank. Tumor dimensions were measured with calipers over 21 days to monitor growth. At the study endpoint, mice were euthanized via cervical dislocation. Tumors were excised, photographed, and weighed. Tumor volume was computed using the formula: Volume (mm³) = 0.5 × length × width².

RT-qPCR

Total RNA was isolated from cultured BC cells using Trizol reagent (Invitrogen) as per the manufacturer’s instructions. RNA concentration and purity (A260/A280 ratio) were assessed by spectrophotometry. Equal amounts of total RNA were reverse transcribed into cDNA employing the M-MLV Reverse Transcriptase cDNA Synthesis Kit (Takara, Kusatsu, Shiga, Japan). Quantitative PCR was executed on an ABI PRISM 7700 platform (Thermo Fisher Scientific, Waltham, MA, USA) using SYBR GreenER qPCR SuperMix (Thermo Fisher Scientific). Gene-specific primers (designed for specificity and commercially synthesized) were utilized. Each 25 µL reaction mixture contained 0.2 µg cDNA, 0.2 µM of each primer, and master mix. The thermal cycling program comprised an initial denaturation at 95°C for 5 min, followed by 40 cycles of 95°C for 15s and 60°C for 45s. Fluorescence was captured during the annealing/extension phase. The 2^−ΔΔCt method was used to calculate relative gene expression, normalized to an internal reference gene. Data analysis was performed with ABI Prism 7700 SDS Software (Thermo Fisher Scientific). Primer sequences are provided in Table 1.

Table 1.

Primer Sequences for RT-qPCR

Gene	Primer Sequence (5’–3’)
CSTF2	Forward: 5’- CAGGGTGGATCGTTCTCTAC −3’
	Reverse: 5’-AACAACAGGTCCAACCTCAGA-3’
THSD1	Forward: 5’-TTCAAGGAGGCCCCCAGATA-3’
	Reverse: 5’-CCACAGTCCCCAAGTTTCTGTAT-3’
PD-L1	Forward: 5’-GCGAAAGCAGAGGAGGA −3’
	Reverse: 5’-TCACAGGCGTCGATGAG −3’
β-actin	Forward: 5’-CTGGAACGGTGAAGGTGACA-3’
	Reverse: 5’-GGGACTTCCTGTAACA ACGCA-3’

EdU Assay

Cellular proliferation was evaluated with a commercial EdU assay kit following the supplier’s protocol. In brief, BC cells were seeded in 24-well plates and grown for 48 h. Cells were then exposed to a 10 μmol/L EdU working solution (Beyotime) for 2 h at 37 °C. After incubation, cells were washed with PBS, fixed with 4% paraformaldehyde for 15 min at room temperature, and permeabilized with 0.5% Triton X-100. Incorporated EdU was detected via a click reaction with the supplied Click Additive Solution, conducted in the dark for 30 min. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI), and fluorescence images were acquired with an Olympus FSX100 microscope (Tokyo, Japan). The proliferation index was finally determined as the ratio of EdU-positive nuclei to total DAPI-stained nuclei, quantified with ImageJ software (National Institutes of Health, Bethesda, MD, USA).

Scratch Wound Healing Assay

The migratory potential of BC cells was tested using a scratch wound healing assay. BC cells were grown to confluence. A uniform scratch was made across the monolayer using a sterile 200 μL pipette tip. After scratching, the medium was replaced with serum-free medium. Control, CSTF2-knockdown, and CSTF2-overexpressing cells were then cultured for 24 h. Wound closure was documented by phase-contrast microscopy at 0 and 24 h. The migration rate was determined using the formula: [(Initial wound area at 0 h − Wound area at 24 h) / Initial wound area at 0 h] × 100%.

Colony Formation Assay

The clonogenic capacity of cells was evaluated using a standard colony formation assay. Cells in the exponential growth phase were detached by trypsin digestion to obtain a single-cell suspension. A predetermined number of cells (500–1000 cells per well) were seeded into 6‑well plates and maintained in culture without disturbance for a period of up to 14 days. After the incubation, the culture medium was carefully removed. Colonies were then washed with PBS, fixed with 4% paraformaldehyde for 15 min, and stained using 0.1% crystal violet solution for 9 min. Following a final wash and air‑drying, colonies containing 50 or more cells were enumerated manually under a light microscope. The resulting colony count per well served as a measure of clonogenic survival.

Migration Assay

To evaluate cell migration, a Transwell chamber-based assay was utilized. The lower wells were filled with a complete medium supplemented with 15% FBS serving as a chemoattractant. BC cells were harvested and adjusted to a concentration of 3 × 10⁴ cells/mL in a serum-free medium before being introduced into the upper chambers. These chambers were equipped with 8.0 µm pore polycarbonate membranes and filled with serum-free medium. After a 48-h incubation period at 37°C with 5% CO₂, non-migratory cells on the upper membrane surface were gently removed using a swab. Cells that had migrated through the membrane were fixed with 4% paraformaldehyde for 10 min, stained with 0.1% crystal violet for 10 min, rinsed with PBS, and air-dried before being mounted. The quantification of migrated cells was conducted by counting cells in five randomly selected microscopic fields per chamber using a light microscope.

Flow Cytometry

Surface expression of PD-L1 on BC cells was analyzed by flow cytometry. Cells were seeded in 6-well plates at a density of 1 × 10⁶ cells per well and grown until they reached about 80% confluence. After harvesting and washing with cold PBS, cells were subjected to antibody staining. Specifically, cells were incubated with a primary antibody against PD-L1 (abcam, ab205921) diluted 1:500 for 1 h at room temperature. Following washes to remove unbound antibody, a fluorochrome-conjugated secondary antibody was applied, and cells were incubated for 30 min in the dark. After a final wash, stained cells were resuspended in 200 μL PBS for analysis. Data were acquired on a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA, USA), collecting at least 10,000 events per sample. The percentage of PD-L1-positive cells was determined by comparison with an isotype-matched control.

RNA Immunoprecipitation (RIP) Assay

An RIP assay was performed to investigate interactions between CSTF2 and target mRNAs. Cells were lysed in RIP lysis buffer (Beyotime) containing protease and RNase inhibitors. Lysates were incubated overnight at 4°C with an anti-CSTF2 antibody or control IgG. Protein A/G magnetic beads (MCE, Monmouth Junction, NJ, USA) were introduced to capture immune complexes. After washing, co-immunoprecipitated RNA was extracted using Trizol reagent and reverse-transcribed into cDNA. Enrichment of THSD1 mRNA in the immunoprecipitates relative to input RNA was quantified by RT-qPCR.

Fluorescence in situ Hybridization (FISH) and Immunofluorescence (IF)

Subcellular localization and co-localization of THSD1 mRNA and CSTF2 protein were analyzed using a combined FISH/IF kit (GenePharma, Suzhou, China). BC cells grown on coverslips were fixed with 4% paraformaldehyde for 10 min and permeabilized with 0.5% Triton X-100 for 5 min at 4°C. After pre-hybridization, cells were hybridized overnight at 37°C with a Cy3-labeled THSD1 probe (2.5 µL, 20 µM; GenePharma). Unbound probe was removed by washing, and nuclei were stained with DAPI. For IF, samples were blocked with 5% BSA for 30 min at 37°C, incubated with an anti-CSTF2 primary antibody (1:250; abcam, ab138486), and then exposed to a fluorophore-conjugated secondary antibody. Images were acquired using a confocal microscope (Olympus) to evaluate co-localization of THSD1 mRNA (red) and CSTF2 protein (green).

mRNA Degradation Assay

An mRNA degradation assay employing actinomycin D was conducted to clarify CSTF2’s effect on THSD1 mRNA stability. BC cells were seeded in 6-well plates at 1 × 10⁶ cells/well. Upon reaching appropriate confluence, transcription was inhibited by adding 5 µM actinomycin D. Cells were collected at 0, 5, 10, and 20 h post-treatment. Total RNA was extracted at each time point using the RNeasy Mini Kit (Qiagen, Hilden, Germany). THSD1 mRNA levels were measured by RT-qPCR and normalized to a reference gene. The mRNA decay rate (half-life) was derived from the slope of a semi-logarithmic plot of remaining mRNA percentage versus time.

Immunohistochemistry (IHC)

IHC staining was performed on formalin-fixed, paraffin-embedded tissue sections from xenograft tumors. After resection, tumor specimens were fixed by immersion in 10% neutral buffered formalin for 24–48 h. Tissues were dehydrated through a graded ethanol series, cleared in xylene, and embedded in paraffin blocks. Sections of 4–5 µm thickness were cut with a microtome, placed on poly-L-lysine-coated glass slides, and deparaffinized and rehydrated. For antigen retrieval, slides were heated in either 10 mM sodium citrate buffer (pH 6.0) or Tris-EDTA buffer (pH 9.0). Subsequent steps included quenching endogenous peroxidase with 3% H₂O₂ for 10 min and blocking nonspecific sites with appropriate normal serum. Sections were incubated overnight at 4°C with primary antibodies: anti-PD-L1 (abcam, ab205921), anti-CD8 (abcam, ab217344), and anti-Ki-67 (abcam, ab21700). Following thorough washing, HRP-conjugated secondary antibodies were applied and incubated for 1 h at room temperature. Signals were visualized using 3,3’-diaminobenzidine as the chromogen, followed by counterstaining with hematoxylin. Finally, stained sections were dehydrated, cleared, and coverslipped. All slides were examined with an Olympus bright-field microscope, and representative fields were captured for analysis.

Bioinformatics Analysis

Publicly available genomic and transcriptomic datasets were analyzed to investigate the role of CSTF2 in BC using several established bioinformatics platforms. Analyses of differential gene expression, patient survival, and correlation patterns based on The Cancer Genome Atlas (TCGA) data (https://portal.gdc.cancer.gov/) were performed using the GEPIA (http://gepia.cancer-pku.cn/) and KM-plotter (http://kmplot.com) online platforms. The immune cell composition within tumors in relation to CSTF2 expression was analyzed using the Clinical Bioinformatics Home platform (https://www.aclbi.com/static/index.html#/). CSTF2‑associated genomic alterations, including copy‑number variations and mutational status, were examined using the cBioPortal for Cancer Genomics (https://www.cbioportal.org/). Furthermore, to elucidate potential regulatory networks, putative CSTF2 target genes were predicted using the ENCORI database (https://rnasysu.com/encori).

Statistical Analysis

Data are expressed as the mean ± standard error of the mean (SEM) from a minimum of three independent experiments. Differences between two experimental groups were evaluated for statistical significance using a two-tailed, unpaired Student’s t-test. Comparisons involving multiple groups were analyzed by one-way or two-way analysis of variance (ANOVA), as appropriate, with subsequent post-hoc testing performed using either Tukey’s or Šidák’s method. A probability (P) value of less than 0.05 was defined as statistically significant. All statistical computations and graphical representations were conducted using GraphPad Prism 10 software.

Results

CSTF2 is Highly Expressed in BC and Significantly Associated with Poor Prognosis

To investigate CSTF2’s potential oncogenic role in BC, its expression was initially assessed using public genomic databases. Analysis of TCGA-BRCA datasets indicated that tumor tissues harbored significantly higher CSTF2 transcript levels relative to paired normal tissues (Figure 1A). Stratification by molecular subtype (Luminal A [LumA], Luminal B [LumB], HER2-enriched [HER2+], Basal-like) revealed CSTF2 expression was most pronounced in the Luminal B subtype (Figure 1B and C), which is frequently linked with poorer outcomes. Survival analysis using KM-plotter and independent datasets (GSE2603, GSE25066, GSE46184) corroborated that elevated CSTF2 expression correlated with diminished overall survival (Figure 1D–F). Genomic analysis via cBioPortal indicated a positive correlation between CSTF2 mRNA abundance and gene copy number (Figure 1G), implying amplification may underlie its overexpression. For experimental corroboration, CSTF2 expression was assessed in cell lines. RT-qPCR and Western blot analyses of the normal mammary epithelial line MCF-10A and a panel of BC lines (MCF7, T47D, MDA-MB-468, MDA-MB-231, HS578T) demonstrated consistently higher CSTF2 mRNA and protein levels in cancer cells compared to MCF-10A (Figure 1H and I).

Figure 1.

CSTF2 is Highly Expressed in BC and Significantly Associated with Poor Prognosis. (A) Analysis of the TCGA-BRCA database shows high expression of CSTF2 in BC. (B) Analysis of CSTF2 expression across the four molecular subtypes of BC (LumA, LumB, HER2+, Basal) in the TCGA-BRCA database. (C) Survival analysis of CSTF2 across the four molecular subtypes (LumA, LumB, HER2+, Basal) in BC from the TCGA-BRCA database. (D–F) Overall survival of BC patients with low vs high CSTF2 expression in the GSE2603, GSE25066, and GSE46184 datasets. (G) Correlation analysis between CSTF2 expression and its copy number variation via the cBioPortal database. (H and I) mRNA and protein expression levels of CSTF2 detected in human mammary epithelial MCF-10A cells and BC cell lines (MCF7, T47D, MDA-MB-468, MDA-MB-231, HS578T). Error bars represent data from at least three independent experiments expressed as mean ± SEM. ** P < 0.01, *** P < 0.001, **** P < 0.0001 vs MCF-10A; ns, no significance. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

Overexpression of CSTF2 Promotes the Proliferation and Migration of BC Cells

To functionally delineate CSTF2’s role in breast carcinogenesis, gain-of-function experiments were undertaken. The T47D cell line, possessing relatively high basal CSTF2 expression, was chosen as the primary model. A CSTF2 overexpression plasmid (pcDNA3.1-CSTF2) was introduced into T47D cells. Successful overexpression was verified by RT-qPCR and Western blot, compared to cells transfected with the empty vector (pcDNA3.1) (Figure 2A and B). The impact of CSTF2 overexpression on proliferation was evaluated via colony formation and EdU assays. Relative to control cells, CSTF2-overexpressing cells displayed a marked increase in colony number and EdU-positive nuclei, signifying enhanced proliferative capacity (Figure 2C and D). Consistent with these findings, Western blot showed CSTF2 overexpression elevated protein levels of Ki-67 and PCNA (Figure 2E). The effect on migration was examined using scratch wound healing assay and Transwell assays. CSTF2-overexpressing cells exhibited accelerated scratch closure and increased migration in Transwell assays (Figure 2F and G). These results indicate that CSTF2 overexpression fosters proliferation and migration in BC cells.

Figure 2.

Overexpression of CSTF2 Promotes the Proliferation and Migration of BC Cells. (A and B) RT-qPCR and Western blot analysis of CSTF2 mRNA and protein expression levels after transfection with the overexpression plasmid. (C and D) Colony formation and EdU assays evaluating the effect of CSTF2 overexpression on cell proliferation. Scale bars, 100 μm. (E) Western blot analysis of the effect of CSTF2 overexpression on Ki-67 and PCNA protein expression levels. (F and G) Cell migration assessed by scratch wound healing and Transwell assays after CSTF2 overexpression. Scale bars, 100 μm. Error bars represent data from at least three independent experiments expressed as mean ± SEM. * P < 0.05, ** P < 0.01, *** P < 0.001 vs pcDNA3.1. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

Knockdown of CSTF2 Inhibits the Proliferation and Migration of BC Cells

To examine the functional consequences of CSTF2 depletion, a loss-of-function approach was employed in T47D cells transfected with two distinct siRNAs targeting CSTF2. Knockdown efficiency was evaluated by RT-qPCR and Western blot. siCSTF2-2 yielded a more substantial reduction in CSTF2 expression than siCSTF2-1 and was thus selected for subsequent experiments (Figure 3A and B). The effect of CSTF2 knockdown on proliferation was assessed by colony formation and EdU assays. CSTF2-knockdown cells showed a significant reduction in colony numbers and EdU-positive nuclei compared to control siRNA-treated cells (Figure 3C and D). Concordantly, Western blot revealed decreased protein levels of Ki-67 and PCNA following CSTF2 knockdown (Figure 3E). Migration was evaluated by scratch wound healing assay and Transwell assays. Wound closure was slower in CSTF2-depleted cells, and Transwell assays demonstrated fewer migrating cells upon CSTF2 knockdown (Figure 3F and G). These findings indicate that silencing CSTF2 compromises the proliferative and migratory abilities of BC cells.

Figure 3.

Knockdown of CSTF2 Inhibits the Proliferation and Migration of BC Cells. (A and B) RT-qPCR and Western blot analysis of CSTF2 mRNA and protein expression levels after siRNA transfection. (C and D) Colony formation and EdU assays evaluating the effect of CSTF2 knockdown on cell proliferation. Scale bars, 100 μm. (E) Western blot analysis of the effect of CSTF2 knockdown on Ki-67 and PCNA protein expression levels. (F and G) Cell migration assessed by scratch wound healing and Transwell assays after CSTF2 knockdown. Scale bars, 100 μm. Error bars represent data from at least three independent experiments expressed as mean ± SEM. * P < 0.05, ** P < 0.01, *** P < 0.001 vs siNC. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

CSTF2 Influences the Immune Evasion Process via PD-L1

Cancer immunoediting, which includes the phases of elimination, equilibrium, and escape,²¹ is fundamental to tumor immune evasion. A bioinformatics analysis of the TCGA-BRCA cohort revealed a correlation between CSTF2 expression levels and the estimated abundance of specific tumor-infiltrating immune cells (Figure 4A). Elevated CSTF2 expression was linked to reduced overall immune cell infiltration within the tumor microenvironment (TME) (Figure 4B). Given the central role of the PD-1/PD-L1 axis, the effect of CSTF2 on PD-L1 was probed. Immunoblot analysis revealed PD-L1 protein levels rose with CSTF2 overexpression and fell with its knockdown in T47D cells (Figure 4C and D). Flow cytometry confirmed that CSTF2 overexpression increased, whereas knockdown decreased, the proportion of PD-L1-positive T47D cells (Figure 4E and F). A rescue experiment was conducted to establish causality: CSTF2 was re-expressed in knockdown cells. Re-expression of CSTF2 restored PD-L1 protein levels, as evidenced by Western blot and flow cytometry (Figure 4G and H), indicating PD-L1 expression is directly modulated by CSTF2. These observations suggest CSTF2 contributes to an immunosuppressive TME, at least partly, by upregulating PD-L1.

Figure 4.

CSTF2 Influences the Immune Evasion Process. (A) Correlation network map between CSTF2 and immune cell infiltration in TCGA-BRCA. (B) Correlation analysis between CSTF2 expression levels and the abundance of immune cells. (C) Western blot analysis of PD-L1 expression after CSTF2 overexpression. (D) Western blot analysis of PD-L1 expression after CSTF2 knockdown. (E and F) Flow cytometry analysis of the positive rate of cell surface PD-L1 after CSTF2 overexpression and knockdown. (G) Western blot analysis of PD-L1 expression after rescue (overexpression of CSTF2 in knockdown cells). (H) Flow cytometry analysis of the cell surface PD-L1 positive rate after rescue (overexpression of CSTF2 in knockdown cells). Error bars represent data from at least three independent experiments expressed as mean ± SEM. * P < 0.05, ** P < 0.01, *** P < 0.001 vs pcDNA3.1, siNC. ### P < 0.001 vs siCSTF2. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

CSTF2 Directly Binds to THSD1 mRNA and Promotes Its Degradation

To elucidate the molecular mechanism whereby CSTF2 contributes to immunoediting, its potential mRNA targets were predicted computationally. Preliminary screening using the ENCORI database, followed by protein-mRNA interaction network analysis with Cytoscape, pinpointed THSD1 mRNA as a leading candidate for direct interaction with CSTF2 (Figure 5A). Analysis of the TCGA-BRCA dataset disclosed a significant inverse correlation between CSTF2 and THSD1 mRNA expression (Figure 5B). To validate this regulatory connection, CSTF2 expression was modulated in BC cells. CSTF2 knockdown elevated THSD1 mRNA and protein levels, while CSTF2 overexpression reduced them (Figure 5C and D). To ascertain if regulation occurs at the level of mRNA stability, RNA decay kinetics were analyzed following actinomycin D treatment. The half-life of THSD1 mRNA was extended upon CSTF2 knockdown and curtailed upon CSTF2 overexpression (Figure 5E and F), indicating CSTF2 promotes THSD1 mRNA degradation. Cellular localization analysis by combined FISH-IF demonstrated co-localization of CSTF2 protein (green) and THSD1 mRNA (red) in the cytoplasm (Figure 5G). RIP assays furnished direct biochemical evidence for a physical interaction, showing enrichment of THSD1 mRNA in anti-CSTF2 immunoprecipitates (Figure 5H). These outcomes demonstrate that CSTF2 directly binds THSD1 mRNA and facilitates its degradation, thereby post-transcriptionally repressing THSD1 expression.

Figure 5.

CSTF2 Directly Binds to THSD1 mRNA and Promotes Its Degradation. (A) Correlation network of CSTF2 predicted by the ENCORI database. (B) Correlation analysis between CSTF2 and THSD1 in TCGA-BRCA. (C) THSD1 protein expression after CSTF2 knockdown and overexpression. (D) THSD1 mRNA levels after CSTF2 knockdown and overexpression. (E and F) qPCR detection of remaining THSD1 mRNA in cells with CSTF2 knockdown or overexpression after actinomycin D treatment. (G) FISH-immunofluorescence assay showing co-localization of CSTF2 and THSD1 in BC cells. (H) RIP-PCR assay detecting the binding between CSTF2 and THSD1 mRNA. Error bars represent data from at least three independent experiments expressed as mean ± SEM. * P < 0.05, ** P < 0.01, *** P < 0.001 vs siNC, pcDNA3.1, IgG. ^## P < 0.01, ^### P < 0.001 vs pcDNA3.1. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

CSTF2 Influences Malignant Growth of BC in vivo

To validate CSTF2’s role in BC progression in vivo, a subcutaneous xenograft mouse model was established. Tumors with stable CSTF2 knockdown (shCSTF2) exhibited significantly reduced final weight and volume after 21 days compared to control tumors expressing a non-targeting shRNA (shNC) (P < 0.01) (Figure 6A–D). Western blot analysis of tumor lysates showed decreased PD-L1 and increased THSD1 protein levels in the shCSTF2 group (Figure 6E). Evaluation of the tumor immune microenvironment indicated heightened intratumoral infiltration of CD8⁺ T cells upon CSTF2 knockdown (Figure 6F). IHC of tumor sections revealed that CSTF2 knockdown attenuated Ki-67 and PD-L1 expression while augmenting CD8⁺ T cell infiltration (Figure 6G). These in vivo findings demonstrate that CSTF2 fosters BC progression by establishing an immunosuppressive niche, characterized by PD-L1 upregulation and compromised CD8⁺ T cell-mediated anti-tumor immunity.

Figure 6.

CSTF2 Influences Malignant Growth of BC in vivo. (A) Representative images of xenograft tumor formation in mice (n = 6 per group). (B) Growth curves and volumes (calculated by the formula 0.5 × length × width²) of xenograft tumors measured weekly for three weeks. (C) Volume measurement of xenograft tumors on day 21. (D) Weight measurement of xenograft tumors. (E) Western blot analysis of CSTF2, THSD1, and PD-L1 expression in tumor tissues from each group. (F) Flow cytometry analysis of the percentage of CD8⁺ T cells in tumor tissues from each group. (G) IHC staining of Ki-67, PD-L1, and CD8 expression in tumor tissues from each group. Error bars represent data from at least three independent experiments expressed as mean ± SEM. ** P < 0.01, *** P < 0.001 vs shNC. Statistical significance was calculated using two-tailed unpaired Student’s t-test (for two groups) or one-way ANOVA followed by Tukey’s test (for more than two groups).

Discussion

This study elucidates the diverse oncogenic pathways mediated by the RNA processing factor CSTF2 in BC. CSTF2 is consistently upregulated across BC molecular subtypes and contributes to disease progression by augmenting proliferation and motility. A specific pathway was delineated wherein CSTF2 imposes post-transcriptional repression on THSD1 by directly binding its mRNA and promoting its degradation. This downregulation of THSD1 results in elevated PD-L1 expression on tumor cells, enabling immune escape. These findings delineate a novel regulatory network in BC and furnish a foundation for exploring CSTF2 as a therapeutic target.

The correlation between high CSTF2 levels and adverse clinical outcomes may be partially attributable to reduced immune cell infiltration. Furthermore, CSTF2 upregulation elevates PD-L1, promoting a transition from an immunogenic “elimination” phase to an immunosuppressive “escape” phase, thereby reinforcing the concept of tumor-intrinsic drivers of immunoediting. Initially characterized as an RBP,²² CSTF2 has documented roles in processes such as reproductive development in rodents²³ and in renal fibrosis, where it promotes a pro-fibrotic TGF-β-CSTF2-FGF2 axis.²⁴ In oncology, elevated CSTF2 expression in HCC correlates with poor prognosis and modulates immune infiltration, potentially through inactivation of the Wnt/β-catenin pathway.²⁵ The CSTF2/DPE2A complex also influences prognosis in HCC.²⁶ In glioblastoma, CSTF2 upregulation promotes tumorigenesis by destabilizing BAD mRNA.²⁷ Similarly, CSTF2 promotes tumor progression in bladder cancer by mediating RAC1 3′-UTR shortening²⁸ and serves as an independent prognostic indicator in oral squamous cell carcinoma.²⁹ These observations across diverse cancer types underscore the broad oncogenic significance of CSTF2. Consistent with previous reports, the present study shows that CSTF2 is significantly overexpressed in BC tissues and is associated with poor prognosis. Knockdown of CSTF2 effectively attenuates BC cell proliferation, invasion, and migration in vitro, and suppresses subcutaneous tumor growth in vivo, collectively highlighting its critical role in BC progression. Notably, CSTF2 levels are inversely correlated with T‑cell abundance, and further experimental evidence indicates that this immunosuppressive effect is likely attributable to CSTF2‑induced upregulation of PD-L1.

A pivotal discovery is the identification of THSD1 as a critical node linking CSTF2 to immune suppression. CSTF2 directly binds THSD1 mRNA and accelerates its decay, as confirmed by stability assays and spatial co-localization. While prior research on THSD1 centered on vascular development and aneurysms,^30,31 its function in BC and its regulatory linkage to CSTF2 are novel. A robust inverse correlation between CSTF2 and THSD1 across subtypes implies diagnostic relevance in PD-L1-positive tumors. THSD1 constrains proliferation and migration, and its loss leads to PD-L1 upregulation and impaired CD8⁺ T cell cytotoxicity, thereby hastening tumor progression. This work expands the functional comprehension of THSD1 beyond a structural capacity, positioning it as a key mediator connecting the extracellular matrix to immune-checkpoint signaling.

Notably, whereas traditional immunoediting research has concentrated on genetic or epigenetic alterations, this study shows that RBPs like CSTF2 constitute a covert regulatory layer for immune checkpoints. CSTF2 indirectly upregulates PD-L1 by degrading THSD1, revealing how tumors co-opt RNA processing machinery to evade immunity. High CSTF2 expression may represent an “immunoediting-adaptive phenotype”, enabling swift transcriptome remodeling under immune pressure. Immunotherapy seeks to mobilize the host’s immune system against cancer. Numerous genes, akin to CSTF2, possess immunomodulatory functions within the TME and are linked to immune evasion and progression.^32,33 For instance, aberrant PD-L1 expression drives immune escape in cervical cancer;³⁴ NEK2 inhibition sensitizes pancreatic tumors to PD-L1 blockade;³⁵ and MLLT6 sustains PD-L1 expression and mediates immune resistance.³⁶ PD-L1 also underpins resistance to chemotherapy and targeted therapies.^36,37 In BC, agents targeting PD-L1 have made significant progress.³⁸ Thus, by clarifying a novel upstream mechanism regulating PD-L1 via the CSTF2-THSD1 axis, this investigation may unveil a new avenue for enhancing immunotherapy efficacy in BC.

Conclusion

This study uncovers the CSTF2–THSD1 axis as a pivotal post-transcriptional regulatory hub in BC immunoediting. CSTF2 is markedly overexpressed in all principal BC subtypes, and its elevated expression correlates with unfavorable patient prognosis. Mechanistically, CSTF2 binds to and degrades THSD1 mRNA, thereby upregulating PD-L1 expression. This action diminishes the cytotoxic capabilities of CD8⁺ T cells, supporting tumor immune evasion and fostering disease progression. Collectively, these findings underscore CSTF2 as a prospective prognostic biomarker and a promising therapeutic target in BC.

Funding Statement

Yongzhou City Guiding Science and Technology Project (No.2023YZ049); Special Research Project of National Health Commission Capacity Building and Continnuing Education Center (No.GWJJZX20251007004); Health Research Project of Hunan Provincial Health Commission (No.20258131); The Guiding Science and Technology Project of Yongzhou City (No.2025YZ006).

Data Available

Data are available from the corresponding author on request.

Ethics Statement

This study was a secondary analysis of publicly available de-identified data from the TCGA and GEO databases, for which informed consent had been obtained by the original data providers. The study protocol was reviewed and approved by the Medical Ethics Committee of The Central Hospital of Yongzhou (Approval No. 2026021401). All procedures complied with the ethical standards of the institutional review board and the Declaration of Helsinki.

All animal experiments complied with the ARRIVE guidelines and were performed in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The experiments and publicly available data were approved by the Medical Ethics Committee of The Central Hospital of Yongzhou (Approval No.2026021401).

Author Contributions

Conceptualization, TangXiaoMan Wang; data curation, Lei Yan and JiaLi Jiang; formal analysis, TangXiaoMan Wang and Jing He; investigation, Jing He; methodology, Lei Yan and JiaLi Jiang; writing—original draft preparation, TangXiaoMan Wang; writing—review and editing, Jing He. All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors have no conflicts of interest to declare for this work.