Targeting PRMT5 to reshape the immune microenvironment and reverse Trop2 ADC resistance in triple-negative breast cancer

Abstract

Background

Triple-negative breast cancer (TNBC) patients often develop resistance to Trop2 antibody-drug conjugate (ADC), and plasmacytoid dendritic cell (pDC) dysfunction is associated with poor prognosis in TNBC, but the underlying mechanism linking these two phenomena remains unclear.

Methods

The immune cell infiltration and prognosis of TNBC were analyzed through the TIMER database. Sacituzumab Govitecan (SG)-resistant MDA-MB231 cells and xenograft tumor models were established, and the function of pDC was regulated with CpG or anti-BDCA-2. The relationship between Protein arginine methyltransferase 5 (PRMT5) and immune cell infiltration was predicted by the Xiantao academic analysis. CCK-8, plate cloning, EdU and other experiments were used to detect cell functions. The functions of pDC were detected by flow cytometry and ELISA experiments. Besides, IP and IB experiments were implemented to detect the ubiquitination level of interferon regulatory factor 3 (IRF3) and the methylation level of midline 2 (MID2), and the protein interactions were verified through the Co-IP experiment.

Results

TNBC tissues resistant to SG showed reduced pDC infiltration and impaired function, associated with downregulation of IRF3. PRMT5 was highly expressed in the cytoplasm of SG-resistant TNBC cells and negatively correlated with pDC cell infiltration. PRMT5 promoted MID2 methylation to induce the ubiquitination and degradation of IRF3. Moreover, silencing PRMT5 activated pDCs, reversed CD8⁺ T cell suppression, and enhanced the sensitivity of drug-resistant TNBC cells to SG in vitro and in vivo.

Conclusion

Targeting PRMT5 reshapes the immune microenvironment by restoring pDC function via the MID2-IRF3 axis, providing a promising strategy to reverse SG resistance in TNBC.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s13402-026-01269-4.

Introduction

Triple-negative breast cancer (TNBC) accounts for approximately 15% to 20% of all breast cancer pathological types. It lacks expression of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 [1]. Chemotherapy is a traditional systemic treatment method, but it has problems such as limited efficacy, obvious toxic side effects, and propensity to develop drug resistance, which lead to poor prognosis for patients with TNBC, with a significantly lower 5-year survival rate compared with other subtypes of breast cancer [2, 3]. In recent years, the emergence of antibody-drug conjugates (ADCs) has brought breakthroughs in the treatment of TNBC. Trop2, as a transmembrane glycoprotein highly expressed in TNBC, has emerged as an important therapeutic target [4]. Trop2 ADCs, such as Sacituzumab Govitecan (SG), specifically bind to Trop2 on the surface of tumor cells and deliver cytotoxic drugs to tumor cells to exert cytotoxic effects, significantly improving the survival benefits of some patients with advanced TNBC [5].

However, the heterogeneity of tumor cells and the complex tumor microenvironment lead to the inevitable development of drug resistance in some patients during treatment, which limits the clinical application of Trop2 ADCs [6, 7]. Currently, most studies on the mechanism of Trop2 ADC resistance focus on the molecular changes of tumor cells themselves, such as downregulation of Trop2 expression, activation of drug efflux pumps, and enhanced DNA damage repair capacity [8, 9]. In the pathological process of TNBC, Trop2 utilizes the tight junction barrier mediated by Claudin 7 to physically impede the infiltration of CD8 + T cells into tumor tissues and inhibit their activation. This makes it difficult for CD8 + T cells to effectively enter the interior of tumor tissues and successfully differentiate into effector cells with cytotoxic functions. Consequently, Trop2 ADCs can only rely on the direct killing effect of their carried payloads, making it hard to eliminate tumor cells and ultimately leading to the development of drug resistance in tumor cells [10]. A growing body of evidence suggests that the remodeling of the tumor immune microenvironment is crucial for the emergence of ADC resistance [11, 12]. The infiltration level and functional state of immune cells in the tumor microenvironment directly affect the efficacy of immunotherapy and ADC treatment. Among them, plasmacytoid dendritic cells (pDCs), as an important innate immune cell, can activate the adaptive immune response by secreting cytokines such as type I interferons, and promote the proliferation of CD8⁺ T cells [13, 14]. This research revealed that the infiltration degree of pDCs in TNBC tissues is closely associated with both the disease-free survival and disease-specific survival of patients, with a high level of pDC infiltration typically signifying a more favorable prognosis [15]. Notably, interferon regulatory factor 3 (IRF3) is a key transcription factor in pDCs that regulates the production of type I interferons. Its activation depends on non-MyD88 pathways such as retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) and integrins [16, 17]. It plays a central role in pDC immune activation, disease onset, and tumor immunity [18]. However, whether the development of TNBC resistance is related to the infiltration of pDCs and the specific regulatory mechanism remain to be further clarified.

Protein arginine methyltransferase 5 (PRMT5), a major enzyme responsible for the monomethylation (MMA) and symmetrical dimethylation (SDMA) of arginine, is involved in various biological processes such as gene transcription regulation, cell cycle progression, and signal pathway transduction [19]. Recent studies have substantiated that PRMT5 is highly expressed across a range of malignant tumors, and it drives tumor progression through regulating the malignant biological functions of tumor cells [20, 21]. Meanwhile, PRMT5 can also reshape the tumor immune microenvironment, inhibit the anti-tumor function of immune cells, and mediate tumor immune escape [22]. Most importantly, a study has elucidated that PRMT5 expression is closely related to pDC functional pathways, which are involved in pDC survival [23].

Multiple converging lines of evidence support PRMT5 as a key candidate linking SG resistance to pDC dysfunction. First, bioinformatics analysis of the TCGA-BRCA cohort in this study revealed a significant negative correlation between PRMT5 expression and pDC infiltration scores (R = -0.323, P < 0.001), suggesting that tumors with high PRMT5 activity harbor fewer functional pDCs. Second, PRMT5 is a documented epigenetic suppressor of type I interferon (IFN) production in innate immune cells: by methylating key components of the cGAS/STING pathway, PRMT5 attenuates cytosolic DNA-sensing and downstream IFN-α/β transcription in antigen-presenting cells (APCs) [22, 24]. Given that pDCs are the principal producers of type I IFN and their function critically depends on IRF3/IRF7-mediated IFN gene transcription, PRMT5 overexpression in the tumor microenvironment is well-positioned to impair pDC activity. Third, PRMT5 has been shown to suppress the maturation and antigen-presenting capacity of dendritic cells and to reshape the immunosuppressive tumor microenvironment by promoting regulatory T cell maintenance and dampening cytotoxic immune responses [25]. Collectively, these observations provide a mechanistic rationale for investigating whether PRMT5 coordinately drives SG resistance and pDC dysfunction in TNBC. However, it has not been reported whether PRMT5 is involved in regulating the resistance of TNBC to Trop2 ADC, or its correlation with the function of pDC.

This study systematically explored the molecular mechanism by which PRMT5 regulates Trop2 ADC resistance in TNBC through clinical tissue sample analysis, cellular functional experiments, and animal model validation. The objective was to provide new therapeutic targets and experimental evidence for reversing the Trop2 ADC resistance in TNBC.

Materials and methods

Bioinformatics analysis

TIMER database (version 2.0, https://compbio.cn/timer2/) was used to analyze the relationship between immune cell infiltration in breast cancer tissues and the overall survival (OS) of patients. This database integrates RNA-seq expression profiles from The Cancer Genome Atlas (TCGA) breast invasive carcinoma (BRCA) cohort, which includes 1,095 primary breast tumor samples and 113 adjacent normal tissue samples. The “Survival” module was employed to estimate the correlation between the infiltration levels of multiple immune cell types (including plasmacytoid dendritic cells (pDCs), CD8⁺ T cells, CD4⁺ T cells, B cells, macrophages, and neutrophils) and patient cumulative survival. Patients were stratified into high and low infiltration groups based on the median (50th percentile) of immune cell infiltration scores, with a follow-up period of 100 months. P-values were calculated using the log-rank test to compare survival differences between groups. The same module was also used to evaluate the correlation between CLEC4C (encoding BDCA-2, a specific pDC marker) expression and OS in the TCGA-BRCA cohort.

The Xiantao Academic platform (https://www.xiantaozi.com/) was used to evaluate the correlation between PRMT5 expression and the infiltration of multiple immune cell types in the TCGA-BRCA cohort. Immune cell infiltration scores were computed using single-sample gene set enrichment analysis (ssGSEA), which quantifies the relative abundance of each immune cell type based on predefined gene signatures; pDC enrichment scores specifically reflect the transcriptional activity of a curated pDC gene set in each tumor sample. Pairwise Spearman rank correlation analysis was performed to assess the association between PRMT5 expression and pDC enrichment scores, and the results were visualized as a bubble plot and scatter plot. P-values < 0.05 were considered statistically significant.

Clinical samples

This study was approved by the Xijing Hospital Ethics Committee, and all patients signed informed consent forms. Inclusion criteria were defined as patients with histologically confirmed TNBC who were treated with SG. The sensitive group (TNBC-S, n = 5) was defined as those with a complete response or partial response to SG treatment that lasted ≥ 6 months. The drug resistance group (TNBC-R, n = 5) was defined as disease progression after SG treatment, including primary drug resistance (no response to treatment) and acquired drug resistance (disease progression after an initial response). All patients had tumor tissue obtained via surgical resection prior to the first SG treatment. For patients with drug resistance who experienced disease progression, tumor tissue was obtained via repeat biopsy after progression. Tissue blocks of 1–2 mm³ were immediately incubated in RNAlater solution at 4℃ for 24 h and subsequently transferred to a -80℃ refrigerator for long-term storage for subsequent molecular biology assays. The remaining tissues were fixed in 10% neutral formalin for 48 h, dehydrated, cleared, and embedded in paraffin to prepare sections for immunohistochemistry (IHC) analysis.

Clinical tissue preprocessing for downstream experiments

Clinical tumor tissues were processed using distinct protocols according to different downstream applications: For IHC: Tissues were fixed in formalin and prepared as formalin-fixed paraffin-embedded (FFPE) sections. For flow cytometry: Fresh tissues were immediately digested into single-cell suspensions. For Western blot: Tissues were stored in RNAlater at − 80 °C and homogenized for protein extraction. For ELISA: Tissues were homogenized and quantified for cytokine detection. Detailed experimental procedures are described in the corresponding method sections below.

IHC assay

The resulting sections were subjected to antigen retrieval with citrate buffer, and nonspecific binding sites were blocked with 5% BSA. Subsequently, the specific marker antibody for pDCs (Anti-CLEC4C antibody, which targets BDCA-2; Cat. No. ab239078, Abcam, Cambridge, UK) was added and incubated overnight at 4℃. The HRP-labeled secondary antibody (Cat. No. ab6721, Abcam) was added the next day and incubated for 1 h at room temperature. Then, DAB chromogen (Cat. No. ab64238, Abcam) was applied, and the staining time was controlled under a microscope. Once the positive signals were clear, the reaction was terminated with distilled water. The sections were counterstained with hematoxylin, dehydrated, cleared, and then mounted with neutral balsam. The number of pDCs was observed under a light microscope.

Determination of the number of pDCs and the positive rate of CD80⁺CD86⁺

The crushed tumor tissue was filtered through a cell strainer, after which centrifugation was performed to collect a single-cell suspension of tumor cells. One million (1 × 10⁶) single cells were taken and resuspended in PBS containing 2% fetal bovine serum (FBS; Thermo Fisher Scientific, Waltham, MA, USA). Non-specific binding was blocked using a CD16 antibody (Cat. No. ab246222, Abcam). Then, the cells were incubated with an Anti-CLEC4C antibody (Cat. No. ab239078, Abcam). After two washes, the cells were resuspended in 200 µL of staining buffer. The percentage of pDCs among the total viable cells was recorded using a flow cytometer.

BDCA-2⁺ pDCs were isolated from the tumor single-cell suspension using a magnetic bead separation kit (Thermo Fisher Scientific). Following isolation, BDCA-2⁺ pDCs were resuspended and cultured in RPMI-1640 medium (Gibco, Thermo Fisher Scientific) supplemented with 10% heat-inactivated FBS (Thermo Fisher Scientific), 1% penicillin/streptomycin (Invitrogen), and recombinant human IL-3 (10 ng/mL; PeproTech, Rocky Hill, NJ, USA) to support pDC survival and minimize spontaneous apoptosis. Cells were maintained at 37 °C in a humidified incubator with 5% CO₂ for no longer than 48 h prior to functional assays. pDC purity was confirmed by flow cytometry (> 85% BDCA-2⁺ cells) before downstream experiments. 1 × 10⁶ sorted BDCA-2⁺ pDCs were resuspended in separation buffer, and non-specific binding was blocked with a CD16 antibody (Cat. No. ab246222, Abcam). A pre-mixed cocktail of fluorescently labeled antibodies (PE-BDCA-2 (Cat. No. ab239078, Abcam) + FITC-CD80 (Cat. No. ab225674, Abcam) + APC-CD86 (Cat. No. ab239075, Abcam)) was then added, and the cells were incubated on ice for 30 min. After staining, the cells were washed and resuspended, and the double-positive rate for CD80⁺CD86⁺ was detected by flow cytometry.

Enzyme-linked immunosorbent assay (ELISA)

The concentrations of IFN-α, IL-6, and TNF-α were respectively determined using the Human/Mouse IFN-α ELISA Kit (Cat. No. PI505, Beyotime, Shanghai, China; Cat. No. E-EL-M3054, Elabscience Biotechnology, Wuhan, China), Human/Mouse IL-6 ELISA Kit (Cat. No. 97068ES/98027ES, Yeasen, Shanghai, China), and Human/Mouse TNF-α ELISA Kit (Cat. No. 97072ES/98029ES, Yeasen).

For ELISA from tissue specimens, tumor tissues were first weighed and then homogenized in ice-cold PBS supplemented with a protease inhibitor cocktail (1:100, Beyotime) at a ratio of 100 mg tissue per 1 mL PBS using a tissue homogenizer (30 Hz, 60 s, 3 cycles). The homogenate was centrifuged at 12,000 × g for 15 min at 4 °C, and the clarified supernatant was collected for ELISA. Total protein concentration in each supernatant was quantified using the BCA Protein Assay Kit (Cat. No. P0010, Beyotime), and cytokine concentrations were normalized to total protein content (expressed as pg/mg protein). For cell culture supernatant ELISA, supernatants were collected after 24 h of indicated treatments and centrifuged at 300 × g for 5 min to remove cell debris prior to assay. All ELISA assays were performed strictly according to the respective manufacturer’s protocols. Briefly, standards and samples (100 µL/well) were added to pre-coated 96-well plates and incubated at 37 °C for 2 h. After washing, detection antibody solution was added and incubated for 1 h at 37 °C, followed by incubation with HRP-conjugated streptavidin for 30 min. Colorimetric development was performed using TMB substrate solution (10 min, room temperature, protected from light), and the reaction was terminated with stop solution. Absorbance was measured at 450 nm (reference wavelength: 570 nm) using a microplate reader. A four-parameter logistic curve was used to calculate sample concentrations from the standard curve.

Western blot

Proteins were extracted using RIPA lysis buffer (Cat. No. P0013B, Beyotime) and quantified with a BCA protein assay kit (Cat. No. P0010, Beyotime). Subsequently, 50 µg of protein was subjected to SDS-PAGE and transferred onto a PVDF membrane. After blocking with 5% non-fat milk, the membrane was incubated with primary antibodies overnight at 4℃. The following day, the secondary antibody (Cat. No. ab6721, Abcam) was added and incubated at room temperature for 1 h. Protein bands were visualized using an ECL kit (Cat. No. P0018M, Beyotime) and quantified using the ImageJ software. Primary antibodies: Anti-IRF3 antibody (1:1000, Cat. No. ab76409, Abcam), Anti-PRMT5 antibody (1:1000, Cat. No. ab109451, Abcam), Anti-GAPDH antibody (1:10000, Cat. No. ab181602, Abcam), Anti-Lamin B1 antibody (0.1 µg/mL, Cat. No. ab16048, Abcam), and Anti-MID2 antibody (1:1000, Cat. No. ab188474, Abcam).

Cell culture and construction of TNBC-resistant cell lines

MDA-MB231 cells (Cat. No. SNL-073, SUNNCELL, Wuhan, China) were cultured in Leibovitz’s L-15 medium (Cat. No. C2786, Beyotime) containing 10% FBS (Thermo Fisher Scientific) and 1% penicillin/ streptomycin (Invitrogen, Carlsbad, CA, USA), maintained at 37℃ in a 5% CO₂ atmosphere.

The drug-resistant cell line (MDA-MB-231/R) was induced and established via the stepwise concentration escalation method. Specifically, MDA-MB-231 cells were inoculated in a culture medium containing a low concentration of SG. After continuous cultivation for 2 to 3 weeks, the concentration of SG was gradually increased, and finally, the cell line that could stably grow in medium with high SG concentrations was selected. The half-maximal inhibitory concentration (IC50) of SG for the parental and resistant cell lines was detected using the CCK-8 kit (Cat. No. C0038, Beyotime).

BT549 cells (catalog number HTB-122, ATCC) were cultured in RPMI-1640 medium containing 10% FBS and 1% penicillin/streptomycin, and the SG-resistant BT549/R cell line was established using the same stepwise concentration escalation method as MDA-MB231/R cells.

Cell apoptosis detection

The assessment of apoptosis rate was achieved through the use of the Annexin V-FITC/PI Apoptosis Detection kit (Cat. No. 40302ES20, Yeasen). A total of 1 × 10⁵ cells were collected and resuspended in 1×Binding Buffer. 5 µL of Annexin V-FITC and 10 µL of propidium iodide staining solution were added sequentially and incubated for 20 min at room temperature in the dark. Finally, apoptosis was assessed by flow cytometry.

Cell proliferation detection

Cell proliferation was detected using the EdU-594 cell proliferation assay kit (Cat. No. C0078S, Beyotime). Cells were seeded in 6-well plates and cultured to logarithmic phase and treated accordingly. A 2× EdU working solution was added and incubated in the incubator for 2 h. Cells were fixed and permeabilized using 4% paraformaldehyde and 0.3% Triton X-100/PBS. 500 µL of the configured Click reaction solution was added to each well and incubated at room temperature in the dark for 30 min. Nuclei were subsequently stained with 1×Hoechst 33,342. Cell proliferation was observed under a fluorescence microscope, and the positive rate of the 594 channel was detected by flow cytometry.

Construction of SG-resistant xenograft tumor model

All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Xijing Hospital and were carried out strictly in compliance with pertinent guidelines. The 4-6-week-old specific pathogen-free (SPF) grade C57BL/6 nude mice (SPF Biotechnology, Beijing, China) were purchased and acclimated for 1 week. MDA-MB231/R cells in logarithmic growth phase were harvested and adjusted to a cell concentration of 5 × 10⁶ cells/ml using PBS. The mice were then divided into 4 groups (n = 5): the Normal group, the SG group (SG therapeutic concentration: 10 mg/kg), the SG + CpG group, and the SG + CpG + anti-BDCA-2 group. 200 µL of the cell suspension was injected subcutaneously into the right back of each nude mouse. The mental state, diet and transplantation site of nude mice were observed daily after inoculation. CpG ODN 2216 (CpG) was a human-specific Toll-like receptor 9 (TLR9) ligand or agonist that induces IFN-α via pDC [26]. In this study, pDCs were activated by injection of CpG (25 µg per mouse; Cat. No. TLRL-2216, Invivo, France). Anti-BDCA-2 monoclonal antibody (anti-BDCA-2; Cat. No. HY-P9990, MedChemExpress, New Jersey, USA) was used to inhibit the activation of pDC [27]. 21 days later, the nude mice were anesthetized, euthanized, and quickly decapitated. The tumor tissues were then removed and weighed, and their volume was measured using the formula: Volume = (L × W²) /2.

The pDCs detected in the tumor microenvironment of these xenograft models are of murine origin, as the implanted MDA-MB231/R cells are human-derived and do not give rise to host immune cells. Murine pDCs were identified using the anti-BDCA-2 (anti-CLEC4C) antibody (Cat. No. ab239078, Abcam), which has been validated for cross-reactivity with murine CLEC4C. To further confirm species-specificity, co-staining with an anti-mouse BST2/mPDCA-1 antibody (Cat. No. 127016, BioLegend, San Diego, CA, USA), a well-established murine-specific pDC marker, was performed, and co-expression of BDCA-2 and mPDCA-1 was confirmed in the tumor-infiltrating pDC population. Although CpG ODN 2216 is designated a class A CpG oligonucleotide with preferential activity on human TLR9, it has been documented to stimulate murine TLR9 via MyD88-dependent signaling pathways and to induce IFN-α secretion from murine pDCs in vivo [28].

4-6-week-old BALB/c mice (SPF Biotechnology) were divided into 4 groups (n = 5): the sh-NC group (MDA-MB231/R cells transfected with sh-NC), the sh-PRMT5-2 group (MDA-MB231/R cells transfected with sh-PRMT5-2), the sh-NC + SG group, and the sh-PRMT5-2 + SG group. Mice were euthanized 21 days after SG treatment. The volume and weight of the removed tumors were also measured. Tumor tissues were prepared into 4 μm sections, deparaffinized, and hydrated for TUNEL and Ki67 staining to assess cell apoptosis and proliferation, respectively.

Co-culture experiments

Human pDCs were isolated from peripheral blood mononuclear cells (PBMCs) derived from healthy donors using a human plasmacytoid dendritic cell isolation kit (Cat. No. 130-097-415, Miltenyi Biotec, Bergisch Gladbach, Germany) according to the manufacturer’s protocol. Isolated pDCs were maintained in RPMI-1640 medium (Gibco, Thermo Fisher Scientific) supplemented with 10% FBS, 1% penicillin/streptomycin, and recombinant human IL-3 (10 ng/mL; PeproTech, Rocky Hill, NJ, USA). pDC purity was confirmed by flow cytometry (> 90% BDCA-2⁺ cells). For co-culture assays, MDA-MB231/R cells (1 × 10⁵) were seeded in the lower chamber of a 24-well Transwell system (0.4 μm pore size; Corning, NY, USA) in Leibovitz’s L-15 medium. Human pDCs (5 × 10⁵) were added to the upper chamber at an effector-to-target ratio of 5:1. The co-cultures were treated with CpG ODN 2216 (5 µg/mL), anti-BDCA-2 (10 µg/mL), or their combination as indicated, and incubated for 48 h at 37 °C with 5% CO₂. Following co-culture, supernatants were harvested from the lower chamber and used for ELISA quantification of IFN-α, IL-6, and TNF-α. pDCs from the upper chamber were collected for flow cytometric analysis of CD80⁺CD86⁺ expression. For the LDH release assay, MDA-MB231/R cells and pDCs were co-cultured in direct contact in 96-well plates. All procedures involving human blood samples were approved by the Xijing Hospital Ethics Committee, and all donors provided written informed consent.

For the three-cell co-culture experiments, human CD8⁺ T cells were isolated from PBMCs of healthy donors by negative magnetic selection using the CD8⁺ T Cell Isolation Kit (Cat. No. 130-096-495, Miltenyi Biotec) according to the manufacturer’s instructions. CD8⁺ T cell purity was confirmed by flow cytometry (> 95% CD3⁺CD8⁺ cells). Isolated CD8⁺ T cells were pre-activated with anti-CD3/CD28 beads (Gibco, Thermo Fisher Scientific) at a bead-to-cell ratio of 1:1 for 48 h in RPMI-1640 medium supplemented with 10% FBS and recombinant human IL-2 (100 IU/mL; PeproTech). For co-culture, MDA-MB231/R cells (1 × 10⁵) were seeded in 24-well plates. Human pDCs (5 × 105) and activated CD8⁺ T cells (5 × 10⁵; effector-to-target ratio = 5:1) were added simultaneously in a 1:1 mixture of Leibovitz’s L-15 and RPMI-1640 medium supplemented with 10% FBS. Where indicated, αCD8a antibody (anti-human CD8a, 10 µg/mL; Cat. No. HY-P9910, MedChemExpress) was added to functionally block CD8⁺ T cell activity. Co-cultures were incubated for 48 h at 37 °C with 5% CO₂. After co-culture, CD8⁺ T cells were harvested for flow cytometric phenotypic analyses (IFN-γ⁺CD8⁺), and supernatants were collected for cytokine quantification.

For intracellular IFN-γ detection in CD8⁺ T cells, cells were stimulated with PMA (50 ng/mL), ionomycin (500 ng/mL), and brefeldin A (10 µg/mL) for 4 h at 37 °C following co-culture. Cells were then surface-stained with anti-CD8-APC antibody (Cat. No. 344722, BioLegend), followed by fixation, permeabilization, and intracellular staining with anti-IFN-γ-PE antibody (Cat. No. 502510, BioLegend). The percentage of IFN-γ⁺CD8⁺ T cells among total CD8⁺ T cells was quantified by flow cytometry.

Lactate dehydrogenase (LDH) release assay

In this study, the killing rate of SG on drug-resistant TNBC cells was evaluated by detecting the release of LDH, using the LDH Cytotoxicity Assay Kit (Cat. No. 40209ES76, Yeasen). MDA-MB231/R cells were seeded in 96-well plates and treated with SG. Subsequently, the supernatant from each well was aspirated and transferred to a new 96-well plate. LDH release reagent was added to each well according to the manufacturer’s instructions of the LDH assay kit. The cells were incubated at room temperature away from light for 30 min. The cell killing rate was calculated by measuring the absorbance at 490 nm using a microplate reader.

Cell transfection

shRNAs targeting PRMT5 (sh-PRMT5-1 and sh-PRMT5-2) were obtained from GeneChem (Shanghai, China) (with sh-NC as the control group). The sequences of PRMT5, IRF3 and MID2 were cloned into pcDNA3.1 plasmids to construct overexpression vectors (OE-PRMT5, OE-IRF3, and OE-MID2) (with the empty pcDNA3.1 plasmid as the control group). The recombinant plasmids were transfected into cells by Lipofectamine 3000 (Invitrogen).

qRT-PCR

Total RNA was extracted by Trizol and reverse transcribed to cDNA. Subsequently, the genes were amplified using SYBR Premix Ex Taq II (TaKaRa, Dalian, China) with specific primers (RiBoBio, Guangzhou, China) on the Real-Time PCR Detection System (Bio-Rad, Shanghai, China). The relative expression level of gene mRNA was obtained by the 2^−ΔΔCt method. The specific primer sequences were as follows: IRF3, Forward (5’-3’): CATGTCCTCCACCAAGTCCT, Reverse (5’-3’): GGCTTGTGATGGTCAAGGTT; GAPDH, Forward (5’-3’) AAGGTGAAGGTCGGAGTCA, Reverse (5’-3’): GGAAGATGGTGATGGGATTT.

Protein stability determination

After the cells underwent corresponding treatments, 100 µg/ml Cycloheximide (CHX; Cat. No. HY-12320, MedChemExpress) was added, and proteins were collected at 0, 15, 30, 45, and 90 min for Western blot analysis.

Detection of IRF3 ubiquitination level

MDA-MB231/R cells were transfected with Myc-IRF3 (Cat. No. V032183, NovoPro, Shanghai, China) + HA-Ub (ubiquitin expression plasmid; Cat. No. V017086, NovoPro) and Flag-PRMT5 (Cat. No. V027109, NovoPro) + Myc-IRF3 + HA-Ub, respectively. Cells were harvested and lysed to extract total proteins. Flag-PRMT5 and Flag-PRMT5 binding proteins were specifically pulled down by immunoprecipitation (IP) after incubation with the Flag-tagged antibody (Cat. No. ab205606, Abcam). After elution, immunoprecipitated protein complexes were obtained, which were subjected to immunoblotting (IB) using the HA-tagged antibody (Cat. No. ab9110, Abcam). Some of the cell lysates were used for Western blot, and verified by Flag (Cat. No. ab205606, Abcam), Myc (Cat. No. ab32072, Abcam), and β-actin (Cat. No. ab8227, Abcam) antibodies.

Determination of MID2 methylation levels

Cells were lysed and cell lysates were incubated with MID2 antibody (Cat. No. ab188474, Abcam) or IgG antibody (the control group; Cat. No. ab172730, Abcam), and then Protein A/G magnetic beads (Beyotime) were added to precipitate the antibody-protein complex. The Input, IgG-precipitated, and MID2-precipitated samples were washed and and IB assays were performed using antibodies against SDMA (Cat. No. 13222, Cell Signaling Technology, Massachusetts, USA), MMA (Cat. No. 8015, Cell Signaling Technology), and MID2 (Cat. No. ab188474, Abcam), respectively.

Cell lysates were incubated with MID2 (Cat. No. ab188474, Abcam) or SYM10 (Cat. No. 07-412, Sigma-Aldrich, St. Louis, MO, USA) antibodies, followed by the addition of protein A/G magnetic beads to capture the antibody-protein complex. After washing the precipitate, the IB assay was performed with SDMA (Cat. No. #13222, Cell Signaling Technology), MID2 (Cat. No. ab188474, Abcam), PRMT5 (Cat. No. ab109451, Abcam), and β-actin (Cat. No. ab8227, Abcam) antibodies.

Co-immunoprecipitation (Co-IP) assay

A small amount of cell lysates were used as the Input samples, and antibodies against PRMT5 (Cat. No. ab109451, Abcam) or MID2 (Cat. No. ab188474, Abcam) were used to incubate the lysates, and the IgG antibody (Cat. No. ab172730, Abcam) served as a negative control. The antibody-protein complexes were captured by protein A/G magnetic beads and washed. Input samples and complexes were subjected to IB using PRMT5 (Cat. No. ab109451, Abcam), MID2 (Cat. No. ab188474, Abcam), or β-actin (Cat. No. ab8227, Abcam) antibodies.

Cells were lysed and the Input group was set at the same time. IgG antibody (Cat. No. ab172730, Abcam), MID2 antibody (Cat. No. ab188474, Abcam) or IRF3 antibody (Cat. No. ab76409, Abcam) were added to the remaining cell lysates and co-incubated. Subsequently, pre-washed protein A/G beads were added. The precipitate was collected by centrifugation and washed. IB assays were performed using IRF3 (Cat. No. ab76409, Abcam) or MID2 (Cat. No. ab188474, Abcam) antibodies.

Statistical analysis

SPSS 26.0 software was used for statistical analysis, and the measurement data were expressed as mean ± standard deviation. The Student’s t-test was used for comparison between two groups, one-way analysis of variance was used for comparison between multiple groups, and the Spearman rank correlation test was used for correlation analysis. P < 0.05 was considered statistically significant.

Results

pDC infiltration was reduced and its function was impaired in SG-resistant TNBC tissues

To explore whether there is an association between pDC infiltration and the prognosis of patients with TNBC, the TIMER database was utilized to analyze the relationship between various types of immune cells in TNBC tissues and the OS of patients. It was found that the OS of TNBC patients with low infiltration of pDCs and CD8⁺ T cells was significantly lower than that of patients with high infiltration (Fig. 1A). BDCA-2, encoded by the CLEC4C gene, is a specific marker for pDCs. Its high expression in TNBC tissues was associated with longer OS in patients (Fig. 1B), suggesting that the level of pDC infiltration was closely related to the prognosis of TNBC patients. The immunohistochemistry results showed that the infiltration of BDCA-2⁺ pDC in the SG-sensitive TNBC tissues (TNBC-S) was significantly higher than that in the resistant tissues (TNBC-R) (Fig. 1C). The detection results of flow cytometry also confirmed this, showing a significant decrease in the proportion of pDCs in TNBC-R (Fig. 1D). Meanwhile, the functional test revealed that the positive rate of CD80⁺CD86⁺ in TNBC-R-derived pDCs was significantly lower than that in sensitive tissues (Fig. 1E). Additionally, the concentrations of IFN-α, IL-6, and TNF-α in the culture supernatant were also notably decreased (Fig. 1F). In addition, Western blot results presented that, compared with the TNBC-S group, IRF3 protein in TNBC-R tissues was obviously downregulated (Fig. 1G). Preliminary studies confirmed that the number and function of pDC were reduced in the tissues of TNBC patients with SG resistance, which might be related to the downregulation of IRF3 protein.

Fig. 1.

Number and function of pDC infiltration in breast cancer tissues of SG resistant patients. (A) Kaplan-Meier survival curves showing the correlation between immune cell infiltration levels and overall survival (OS) in TNBC patients analyzed using the TIMER database. (B) Kaplan-Meier survival curve demonstrating the association between CLEC4C (BDCA-2) mRNA expression and OS in the TCGA-BRCA cohort. (C-G) Comparative analysis between SG-sensitive (TNBC-S, n = 5) and SG-resistant (TNBC-R, n = 5) human TNBC tissues. (C) Immunohistochemical staining and quantitative analysis of BDCA-2 (pDC-specific marker) expression in tumor tissues. Scale bar = 50 μm. (D) Flow cytometric quantification of the percentage of BDCA-2⁺ pDCs among total viable cells in tumor single-cell suspensions. (E) Flow cytometric analysis of the percentage of CD80⁺CD86⁺ activated pDCs in isolated BDCA-2⁺ cell populations. (F) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in culture supernatants of isolated pDCs. (G) Western blot analysis and quantification of IRF3 protein expression in tumor tissues. n = 5. Data are presented as mean ± SD. **P < 0.01, ***P < 0.001 by unpaired Student’s t-test

Activation of pDC enhanced the killing effect of SG on SG-resistant TNBC cells

The purpose of this part was to verify whether restoring the function of pDC could improve the sensitivity of drug-resistant TNBC cells to SG. By gradually increasing the concentration of SG, a drug-resistant cell line of MDA-MB231, designated as MDA-MB231/R was established. The CCK-8 results showed that the SG IC₅₀ value of MDA-MB231/R cells was significantly higher than that of the parental MDA-MB231 cells. The IC50 of MDA-MB231 was 20 ng/ml, and the IC50 of MDA-MB231/R was 80 ng/ml (Fig. 2A). The apoptosis rate of MDA-MB231/R cells after SG treatment was evidently decreased via flow cytometry (Fig. 2B). Subsequently, the MDA-MB231/R cell line was used to construct a xenograft tumor model in nude mice. CpG was used to activate pDC, and anti-BDCA-2 was used to inhibit pDC maturation. As shown in Fig. 2C-E, compared with the SG group, the tumor volume and weight in the SG + CpG group were significantly reduced, and anti-BDCA-2 could partially reverse the inhibitory effect of CpG on the tumor volume and weight. Flow cytometry results showed that the positive rate of CD80⁺CD86⁺ of pDC in the SG + CpG group was significantly higher than that in the SG group (Fig. 2F). Meanwhile, the concentrations of IFN-α, IL-6, and TNF-α in tumor tissue homogenates were increased (Fig. 2G). Similarly, these changes were partially abolished by the addition of anti-BDCA-2. Next, MDA-MB231/R cells and pDC were co-cultured in vitro. LDH release assay showed that the presence of CpG increased the killing rate of SG on MDA-MB231/R cells (Fig. 2H). The positive rate of CD80⁺CD86⁺ on pDCs (Fig. 2I) and the concentration of IFN-α in the supernatant (Fig. 2J) were all elevated. However, after the addition of anti-BDCA-2, the aforementioned effect was significantly reversed. The above in vivo and in vitro experimental results all indicated that activating pDC could enhance the inhibitory effect of SG on TNBC tumors and increase the sensitivity of TNBC cells to SG treatment.

Fig. 2.

Activation of pDC can enhance the killing effect of SG on TNBC drug-resistant cells. (A) CCK-8 assay determination of the half-maximal inhibitory concentration (IC50) of SG in parental MDA-MB231 and SG-resistant MDA-MB231/R cells. (B) Flow cytometric analysis of apoptosis rates in MDA-MB231 and MDA-MB231/R cells treated with 80 ng/ml SG for 48 h. (C-G) In vivo efficacy study in BALB/c nude mouse xenograft models (n = 5 per group). Mice were treated with vehicle (Normal), SG (10 mg/kg), SG + CpG ODN 2216 (25 µg/mouse), or SG + CpG + anti-BDCA-2 monoclonal antibody (10 mg/kg). (C) Representative images of excised tumors at day 21 post-treatment. (D) Tumor volume growth curves measured every 3 days. (E) Final tumor weights at day 21. (F) Flow cytometric quantification of CD80⁺CD86⁺ activated pDCs in tumor tissues. (G) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in tumor tissue homogenates. (H-J) In vitro co-culture experiments of MDA-MB231/R cells and human peripheral blood-derived pDCs (n = 3 independent experiments). (H) LDH release assay evaluating SG-induced cytotoxicity against MDA-MB231/R cells. (I) Flow cytometric analysis of CD80⁺CD86⁺ pDC activation. (J) ELISA measurement of IFN-α concentrations in co-culture supernatants. n = 5. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ns: not significant by one-way ANOVA followed by Tukey’s post-hoc test

PRMT5 was highly expressed in SG-resistant TNBC cells and was negatively correlated with pDC and CD8⁺ T cell infiltration

To explore the expression of PRMT5 in different subcellular structures, GAPDH and Lamin B1 were selected as cytoplasmic and nuclear reference proteins, respectively. The expression of PRMT5 protein in MDA-MB231 and MDA-MB231/R cells was detected by Western blot, showing that PRMT5 protein was significantly decreased in the nucleus and significantly upregulated in the cytoplasm in MDA-MB231/R cells (Fig. 3A), which suggested that PRMT5 was transferred from the nucleus to the cytoplasm during the process of the development of cell resistance. The heatmap of Xiantao Academic platform analysis in Fig. 3B demonstrated that PRMT5 expression in TNBC tissues was negatively correlated with the level of pDC infiltration. Spearman correlation analysis confirmed a significant negative correlation between PRMT5 expression and the enrichment score of pDCs (Fig. 3C). In addition, the box plot further showed that pDC enrichment scores in the high PRMT5 expression group were significantly lower than those in the low expression group (Fig. 3D). Collectively, this demonstrated that high expression of PRMT5 inhibited the infiltration of pDCs in the tumor microenvironment.

Fig. 3.

PRMT5 is upregulated in the cytoplasm of TNBC SG-resistant cells and negatively correlated with pDC infiltration. (A) Western blot analysis of PRMT5 protein expression in nuclear and cytoplasmic fractions of MDA-MB231 and MDA-MB231/R cells. GAPDH was used as a cytoplasmic loading control, and Lamin B1 as a nuclear loading control. n = 3. (B) Heatmap showing the correlation between PRMT5 expression and infiltration levels of multiple immune cell types in the TCGA-BRCA cohort analyzed using the Xiantao Academic platform. (C) Spearman rank correlation analysis demonstrating a significant negative correlation between PRMT5 expression and pDC enrichment score (R = -0.323, P < 0.001). (D) Box plot comparing pDC enrichment scores between PRMT5 high-expression and low-expression groups in the TCGA-BRCA cohort. Data are presented as mean ± SD. ***P < 0.001 by unpaired Student’s t-test

Knockdown of PRMT5 could enhance the sensitivity of TNBC drug-resistant cells to SG

Given that PRMT5 was predicted to be highly expressed in TNBC drug-resistant cells in the early stage, PRMT5 in MDA-MB-231/R cells was silenced to observe the impact of SG treatment on cellular biological functions, thereby validating this conclusion. Western blot assay for the PRMT5 protein confirmed that the knockdown efficiency of sh-PRMT5-1 and sh-PRMT5-2 was excellent (Fig. 4A). CCK-8 results showed that the viability of MDA-MB231/R cells was weakened after silencing PRMT5 (Fig. 4B). Subsequently, SG treatment was administered. The results of the plate cloning and EdU experiments showed that after silencing PRMT5, the inhibitory effect of SG treatment on the proliferation of MDA-MB231/R cells was enhanced (Fig. 4C and D). Flow cytometry analysis confirmed that the apoptosis rate of the cells increased after SG treatment (Fig. 4E). As expected, reducing PRMT5 led to a significant increase in the killing rate of SG against MDA-MB231/R cells (Fig. 4F). Furthermore, the positive rate of CD80⁺CD86⁺ cells in pDCs in the sh-PRMT5-1/-2 group was elevated (Fig. 4G), and the concentrations of IFN-α, IL-6, and TNF-α in the supernatant were also significantly increased (Fig. 4H). In summary, these findings further suggested that inhibition of PRMT5 expression was able to enhance the sensitivity of TNBC drug-resistant cells to SG.

Fig. 4.

PRMT5 knockdown promotes pDC activation and enhances the sensitivity of drug-resistant TNBC cells to SG. (A) Western blot verification of PRMT5 knockdown efficiency in MDA-MB231/R cells transfected with sh-NC, sh-PRMT5-1, or sh-PRMT5-2. (B) CCK-8 assay measuring cell viability of MDA-MB231/R cells 72 h after PRMT5 knockdown. (C-H) MDA-MB231/R cells were transfected with sh-NC, sh-PRMT5-1, or sh-PRMT5-2, then treated with 80 ng/ml SG for 48 h (n = 3 independent experiments). (C) Colony formation assay evaluating cell proliferation capacity. (D) EdU assay quantifying the percentage of proliferating cells. Scale bar = 50 μm. (E) Flow cytometric analysis of apoptosis rates. (F) LDH release assay measuring SG-induced cytotoxicity. (G) Flow cytometric analysis of CD80⁺CD86⁺ activated pDCs in co-culture systems. (H) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in co-culture supernatants. n = 3. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001 by one-way ANOVA followed by Tukey’s post-hoc test

Silencing PRMT5 reversed the suppressive state of CD8⁺ T cells by stimulating pDC activation, enhancing the sensitivity of TNBC drug-resistant cells to SG

To explore whether the regulatory effect of PRMT5 on the drug sensitivity of TNBC drug-resistant cells involves immune cell infiltration, MDA-MB231/R cells with PRMT5 silencing were co-cultured with pDCs and CD8⁺ T cells, and the activation of pDCs and CD8⁺ T cells was inhibited to observe the changes in the function of MDA-MB231/R cells. Western blot analysis showed that the protein level of PRMT5 was downregulated in the sh-PRMT5-2 group, the sh-PRMT5-2 + anti-BDCA-2 group, and the sh-PRMT5-2 + αCD8a group (Fig. 5A). At the same time, when the activation of pDCs and CD8⁺ T cells was inhibited, the viability (Fig. 5B) and proliferation (Fig. 5C and D) of MDA-MB231/R cells stably transfected with sh-PRMT5 were enhanced, while the apoptosis rate was decreased (Fig. 5E). And the killing effect of SG on the MDA-MB231/R cells in the sh-PRMT5-2 + anti-BDCA-2 group and the sh-PRMT5-2 + αCD8a group was also suppressed (Fig. 5F). However, after anti-BDCA-2 treatment, the positive rate of CD80⁺CD86⁺ in pDCs (Fig. 5G)—which had originally increased due to PRMT5 knockdown—and the concentration of cytokines (IFN-α, IL-6, and TNF-α) both showed a decreasing trend (Fig. 5H). Nevertheless, the presence of αCD8a did not exert a significant impact on either of them. This might be because αCD8a only blocks the function of CD8⁺ T cells and cannot reverse the already activated state of pDCs, thus having a relatively weak overall effect on cytokine secretion. Furthermore, flow cytometric intracellular staining revealed that the proportion of IFN-γ⁺CD8⁺ T cells in the sh-PRMT5-2 group was significantly higher than that in the sh-NC group, indicating enhanced effector function of CD8⁺ T cells following PRMT5 knockdown (Fig. 5I). This increase was substantially reversed by anti-BDCA-2 treatment (blocking pDC activation), and αCD8a treatment (directly blocking CD8⁺ T cell function) significantly reduced the proportion of IFN-γ⁺CD8⁺ T cells compared with the sh-PRMT5-2 group. These results confirm that the IFN-γ-producing effector function of CD8⁺ T cells is dependent on pDC activation, and that functional blockade of CD8⁺ T cells directly abrogates this effector phenotype, further validating the PRMT5→pDC→CD8⁺ T cell immune axis.

Fig. 5.

Silencing PRMT5 reversed the suppressive state of CD8⁺ T cells by stimulating pDC activation, enhancing the sensitivity of TNBC drug-resistant cells to SG. MDA-MB231/R cells were transfected with sh-NC or sh-PRMT5-2, then co-cultured with pDCs and CD8 + T cells. Where indicated, anti-BDCA-2 antibody (to inhibit pDC activation) or anti-CD8a antibody (to block CD8⁺ T cell function) was added (n = 3 independent experiments). (A) Western blot analysis of PRMT5 protein expression in MDA-MB231/R cells. (B) CCK-8 assay measuring cell viability 72 h after treatment. (C) Colony formation assay evaluating cell proliferation capacity. (D) EdU assay quantifying the percentage of proliferating cells. Scale bar = 50 μm. (E) Flow cytometric analysis of apoptosis rates. (F) LDH release assay measuring SG-induced cytotoxicity. (G) Flow cytometric analysis of CD80⁺CD86⁺ activated pDCs. (H) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in culture supernatants. (I) Flow cytometric quantification of the percentage of IFN-γ⁺ cells among CD8⁺ T cells following 48 h co-culture and 4 h PMA/ionomycin restimulation. n = 3. Data are presented as mean ± SD. * P < 0.05, ** P < 0.01; *** P < 0.001, ns: not significant by one-way ANOVA followed by Tukey’s post-hoc test

PRMT5 promotes the inhibition of pDC activation and SG resistance in TNBC cells by downregulating IRF3

To further clarify the molecular mechanism of PRMT5-induced inhibition of pDC activation and SG resistance, we focused on the downstream factors of PRMT5. The scatter plot of Spearman correlation analysis showed that IRF3 expression was negatively correlated with PRMT5 in TNBC (Fig. 6A). Western blot analysis showed that the expression of IRF3 was downregulated in MDA-MB231/R cells (Fig. 6B), and the protein level of IRF3 was significantly increased after PRMT5 knockdown (Fig. 6C). Overexpression of PRMT5 downregulated the protein level of IRF3, but overexpression of IRF3 did not affect the expression of PRMT5. After overexpression of IRF3 and introduction of OE-PRMT5, the protein level of IRF3 was decreased again (Fig. 6D). Cell function assays revealed that after overexpression of PRMT5, the viability and proliferative capacity of MDA-MB231/R cells were enhanced (Fig. 6E-G), the apoptosis rate was decreased (Fig. 6H), and the killing rate of SG on cells was also notably decreased (Fig. 6I). Meanwhile, the positive rate of CD80⁺CD86⁺ in pDCs (Fig. 6J) and the concentration of cytokines (IFN-α, IL-6, and TNF-α) in the supernatant were both decreased (Fig. 6K). Surprisingly, overexpression of IRF3 could obviously abolish the effects of OE-PRMT5 on MDA-MB231/R cell function, pDC activation, and cytokine secretion. More importantly, when OE-PRMT5 was transfected under the condition of overexpressing IRF3, the effects of overexpressing IRF3 were partially reversed. Therefore, it was confirmed that the regulation of PRMT5 on TNBC cell resistance and pDC cell activation depended on IRF3.

Fig. 6.

PRMT5 promotes the SG resistance of TNBC cells by downregulating IRF3. (A) Spearman rank correlation analysis showing a significant negative correlation between PRMT5 and IRF3 expression in the TCGA-BRCA cohort (R = -0.209, P < 0.001). (B) Western blot analysis of IRF3 protein expression in MDA-MB231 and MDA-MB231/R cells. (C) Western blot analysis of IRF3 protein expression in MDA-MB231/R cells after PRMT5 knockdown. (D-K) MDA-MB231/R cells were transfected with empty vector, OE-PRMT5, OE-IRF3, or OE-PRMT5 + OE-IRF3 (n = 3 independent experiments). (D) Western blot analysis of PRMT5 and IRF3 protein expression. (E) CCK-8 assay measuring cell viability 72 h after transfection. (F) Colony formation assay evaluating cell proliferation capacity. (G) EdU assay quantifying the percentage of proliferating cells. Scale bar = 50 μm. (H) Flow cytometric analysis of apoptosis rates. (I) LDH release assay measuring SG-induced cytotoxicity. (J) Flow cytometric analysis of CD80⁺CD86⁺ activated pDCs in co-culture systems. (K) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in co-culture supernatants. n = 3. Data are presented as mean ± SD. * P < 0.05, ** P < 0.01; *** P < 0.001, ns: not significant by one-way ANOVA followed by Tukey’s post-hoc test

PRMT5 catalyzes SDMA methylation modification on MID2, thereby promoting the ubiquitination and degradation of IRF3

This part of the experiment was designed to further determine the regulatory mechanism of PRMT5 on the expression of IRF3. The qRT-PCR experiment confirmed that overexpression or silencing of PRMT5 did not have a significant effect on the IRF3 mRNA expression (Fig. 7A and B). However, through Western blot analysis, it was found that overexpression of PRMT5 could promote the degradation of IRF3, and the proteasome inhibitor MG132 could block this effect (Fig. 7C). The CHX experiment demonstrated that overexpression of PRMT5 significantly shortened the half-life of IRF3 (Fig. 7D). Additionally, the ubiquitination detection results indicated that overexpression of PRMT5 could remarkably increase the ubiquitination level of IRF3 (Fig. 7E). The result in Fig. 8A showed that specific SDMA and MMA modifications were enriched on MID2 protein (IP: MID2 lane), whereas no methylation signal was detected in the IgG negative control lane. As PRMT5-catalyzed arginine methylation is a substoichiometric and dynamic modification, only a fraction of MID2 was modified, leading to a weaker signal compared with the Input lane. The enrichment of SDMA and MMA signals in the MID2-IP lane relative to the IgG negative control (which shows no signal) confirms the specificity of MID2 methylation. The signal intensity in the MID2-IP lane is lower than that in the Input, which is expected, as PRMT5-mediated methylation is a dynamic, partial modification affecting only a fraction of total MID2 protein at any given time — a common characteristic of enzymatic arginine methylation marks. Subsequently, the interaction between PRMT5 and MID2 was confirmed by immunoprecipitation assay (Fig. 8B). Using the SYM10 antibody for methylation detection, we confirmed that SDMA modification existed on MID2 (Fig. 8C). Furthermore, by knocking down PRMT5, we further illustrated that the SDMA modification of MID2 was dependent on the presence of PRMT5 (Fig. 8D). The increased level of SDMA modification on MID2 after re-introduction of PRMT5 further supported the above conclusion (Fig. 8E). In Fig. 9A and B, overexpression or knockdown of MID2 did not affect the mRNA level of IRF3, but the protein level of IRF3 was negatively correlated with the MID2 protein level. Meanwhile, the interaction between MID2 and IRF3 was determined via Co-IP assay (Fig. 9C). Further experiments demonstrated that MG132 could curb the negative regulatory effect of overexpressed MID2 on the IRF3 protein (Fig. 9D). Moreover, overexpressing MID2 could also accelerate the degradation rate of the IRF3 protein (Fig. 9E). Furthermore, ubiquitination detection assay revealed that overexpression of MID2 notably enhanced the ubiquitination level of IRF3, and this effect was further strengthened in the presence of the proteasome inhibitor MG132, whereas knockdown of MID2 exerted the opposite effect (Fig. 9F). Additionally, the Western blot analysis in Fig. 9G showed that after silencing PRMT5, the protein level of MID2 was decreased, while the protein level of IRF3 was increased. Overexpression of MID2 could reverse the stabilizing effect of PRMT5 knockdown on IRF3, causing the protein level of IRF3 to decrease again. In summary, these results collectively indicated that PRMT5 induced the SDMA modification of MID2, thereby driving the ubiquitination and degradation of IRF3.

Fig. 7.

PRMT5 downregulates IRF3 expression by promoting its proteasomal degradation in SG-resistant TNBC cells. (A, B) qRT-PCR analysis of IRF3 mRNA levels in MDA-MB231/R cells with PRMT5 overexpression (A) or knockdown (B). n = 3, ns: not significant. (C) Western blot analysis of IRF3 expression in MDA-MB231/R cells with OE-PRMT5 or vector control, with or without MG132 (10 µM, 6 h) treatment. GAPDH was used as a loading control. (D) CHX chase assay showing IRF3 protein stability in MDA-MB231/R cells with OE-PRMT5 or vector control. Cells were treated with CHX (100 µg/mL) for the indicated times, and IRF3 levels were detected by Western blot (left). Relative band intensities were quantified (right, n = 3). (E) Co-IP assay detecting the effect of PRMT5 on IRF3 ubiquitination. MDA-MB231/R cells were transfected with Flag-PRMT5, Myc-IRF3, and HA-Ub plasmids. Cell lysates were immunoprecipitated with anti-Flag antibody, followed by immunoblotting with anti-HA, anti-Flag, anti-Myc, and anti-β-actin antibodies. Input: 5% total cell lysate; IgG: negative control. Data are presented as mean ± SD. ***P < 0.001, ns: not significant by one-way ANOVA followed by Tukey’s post-hoc test

Fig. 8.

PRMT5 physically interacts with MID2 and mediates its SDMA/MMA methylation modifications. (A) IP assay detecting SDMA and MMA modifications on endogenous MID2 protein. Cell lysates were immunoprecipitated with anti-MID2 antibody or IgG, followed by immunoblotting with anti-SDMA, anti-MMA, and anti-MID2 antibodies. Input: 5% total cell lysate. (B) Co-IP assay verifying the interaction between endogenous PRMT5 and MID2. Cell lysates were immunoprecipitated with anti-PRMT5 antibody or IgG, followed by immunoblotting with anti-MID2 and anti-PRMT5 antibodies. Input: 5% total cell lysate. (C) Independent validation of MID2 SDMA modification using anti-SDMA (SYM10) antibody and anti-MID2 antibody. Input: 5% total cell lysate; IgG: negative control. (D) Western blot analysis showing that MID2 SDMA modification is dependent on PRMT5. (E) IP assay showing that PRMT5 overexpression enhances MID2 SDMA modification.β-actin was used as a loading control for Input samples. No specific signals were detected in IgG control lanes, confirming the specificity of IP enrichment

Fig. 9.

MID2 promotes IRF3 ubiquitination and degradation, and the PRMT5-MID2-IRF3 axis is verified by rescue assays. (A, B) qRT-PCR and Western blot analysis of IRF3 expression in MDA-MB231/R cells with MID2 overexpression or knockdown. n = 3, ***P < 0.001, ns: not significant. (C) Reciprocal Co-IP assays confirming the physical interaction between MID2 and IRF3. Input: 5% total cell lysate; IgG: negative control. (D) Western blot analysis showing that MG132 blocks MID2-mediated downregulation of IRF3. (E) CHX chase assay showing that MID2 overexpression shortens IRF3 protein half-life. (F) IP assay detecting IRF3 ubiquitination in cells with MID2 overexpression or knockdown, with or without MG132 treatment. (G) Rescue assay showing that MID2 overexpression reverses PRMT5 knockdown-mediated IRF3 upregulation. GAPDH/β-actin was used as a loading control. Data are presented as mean ± SD. ***P < 0.001 by one-way ANOVA followed by Tukey’s post-hoc test

Targeting PRMT5 enhanced the sensitivity of TNBC drug-resistant cells to SG in vivo

A mouse xenograft tumor model was established using MDA-MB231/R cells to investigate the regulatory role of PRMT5 in vivo on the drug resistance of TNBC cells. The mice were randomly divided into 4 groups: the sh-NC group, the sh-PRMT5-2 group, the sh-NC + SG group, and the sh-PRMT5-2 + SG group. As shown in Fig. 10A, compared with the sh-NC group, the volume and weight of tumors in the sh-PRMT5-2 group were obviously reduced. After treating the mice in the sh-PRMT5-2 group with SG, the inhibitory effect of SG on the growth of mouse tumors was more significant. Subsequently, TUNEL staining and Ki67 staining were performed on the tumor tissue sections. It was found that the apoptotic cells in the sh-PRMT5-2 group were significantly more than those in the sh-NC group; after being combined with SG treatment, the number of apoptotic cells was further increased. Quantitative analysis showed that the mean density of TUNEL staining in the sh-PRMT5-2 group and the sh-PRMT5-2 + SG group was significantly higher than that in the control group, further confirming the above result. The proportion of Ki67-positive cells in the sh-PRMT5-2 group was significantly lower than that in the sh-NC group; when combined with SG treatment, the inhibitory effect on proliferation was more obvious (Fig. 10B-E). The percentage of CD80 + CD86+ cells in pDCs was significantly elevated in the sh-PRMT5-2 group and the sh-PRMT5-2 + SG group, further supporting the role of PRMT5 gene silencing in the activation of pDC (Fig. 10F and G). Additionally, the levels of IFN-α, IL-6, and TNF-α in the sh-PRMT5-2 group were significantly higher than those in the sh-NC group. The SG treatment also increased the secretion of these cytokines (Fig. 10H). These data validated the tumor-promoting effect of PRMT5 at the in vivo level, suggesting that knocking down PRMT5 could stimulate the activation of pDC and weaken the SG resistance of TNBC.

Fig. 10.

Targeting PRMT5 can enhance the sensitivity of TNBC cells to SG in vivo. C57BL/6 nude mice were subcutaneously injected with MDA-MB231/R cells transfected with sh-NC or sh-PRMT5-2, then treated with vehicle or SG (10 mg/kg) (n = 5 per group). (A) Tumor volume growth curves and final tumor weights at day 21 post-treatment. (B) TUNEL staining detecting apoptotic cells in tumor tissues. Scale bar = 10 μm. (C) Ki67 staining detecting proliferating cells in tumor tissues. Scale bar = 10 μm. (D) Quantitative analysis of the mean density of TUNEL-positive cells. (E) Quantitative analysis of the percentage of Ki67-positive cells. (F-G) Flow cytometric quantification of CD80⁺CD86⁺ activated pDCs in tumor tissues. (H) ELISA measurement of IFN-α, IL-6, and TNF-α concentrations in tumor tissue homogenates. n = 5. Data are presented as mean ± SD. * P < 0.05, ** P < 0.01; *** P < 0.001, ns: not significant by one-way ANOVA followed by Tukey’s post-hoc test

Validation of the PRMT5-MID2-IRF3 axis in BT549/R SG-resistant TNBC cells

To confirm that the PRMT5–MID2–IRF3 axis and its regulation of SG resistance are not restricted to MDA-MB231/R cells, we established a second SG-resistant TNBC cell line BT549/R via stepwise SG induction. The IC₅₀ of SG in BT549/R cells was markedly higher than that in parental BT549 cells (Fig. S1A). Consistent with MDA-MB231/R cells, BT549/R cells exhibited enhanced cytoplasmic localization of PRMT5, and decreased IRF3 protein level (Fig. S1B). Knockdown of PRMT5 in BT549/R cells significantly restored IRF3 expression (Fig. S1C), augmented SG-induced cytotoxicity (Fig. S1D), and restored the activation of pDCs (elevated CD80⁺CD86⁺ proportion; Fig. S1E). These results demonstrate that the regulatory mechanism revealed in this study is universally applicable in multiple SG-resistant TNBC cell models.

Discussion

Trop2 ADC, representing a notable milestone in recent TNBC treatment advancements, has markedly enhanced the survival outlook for patients with advanced-stage disease. However, the emergence of drug resistance remains a major challenge in clinical practice. This study is the first to reveal that targeting PRMT5 can restore the function of pDC by regulating the MID2-IRF3 axis, to reshape the immune microenvironment of TNBC, and thereby reverse the drug resistance of tumor cells to the Trop2 ADC (SG), providing a new target and experimental basis for overcoming the treatment bottleneck of TNBC in clinical practice.

A previous study has shown that high pDC levels in TNBC can enrich inflammatory and immune-related gene sets, including the IFN-γ signaling pathway. The infiltration level is closely related to the survival rate of TNBC patients [29]. The present study found that the level of pDC infiltration was significantly reduced and the function was severely impaired in SG-resistant TNBC tissues, accompanied by the down-regulation of IRF3 protein expression. Impaired pDC function was confirmed to be involved in tumor immune escape [30]. At the same time, other studies have shown that promoting the activation of DCs can enhance anti-tumor immunity and inhibit the formation of immunosuppressive tumor microenvironment in TNBC [31–33]. Further experiments in vitro and in vivo confirmed that activation of pDC could significantly enhance the killing effect of SG on drug-resistant cells, while blocking pDC function could weaken the efficacy of SG, suggesting that pDC-mediated anti-tumor immune response is an important mechanism for Trop2 ADC to exert its therapeutic effect. PRMT5, as a crucial epigenetic regulatory factor, has been extensively studied and confirmed to promote the development of cancer and participate in the generation of drug resistance in TNBC [34, 35]. Therefore, it has become a new potential target in the field of TNBC treatment [36, 37]. Furthermore, the research conducted by Mi et al. has confirmed that inhibiting PRMT5 can stimulate the activation of DCs [38]. Herein, it was found that PRMT5 was highly expressed in the cytoplasm of SG-resistant TNBC cells, and was negatively correlated with the infiltration level of pDCs. Inhibition of PRMT5 reversed CD8⁺ T cell quiescence by promoting pDC activation, which was similar to previous results [39]. The mechanistic connection between pDC activation and restored SG sensitivity can be understood through the following cascade: PRMT5-mediated SDMA methylation of MID2 promotes MID2-driven ubiquitination and proteasomal degradation of IRF3 in tumor-associated pDCs. As IRF3 is the primary transcription factor driving type I IFN (IFN-α/β) gene expression in pDCs [40], its loss impairs the ability of pDCs to produce IFN-α in the tumor microenvironment. Tumor-derived type I IFN directly enhances the effector function of CD8⁺ T cells — including IFN-γ production and cytolytic activity — and promotes tumor immune surveillance [41]. Our data confirm this axis: CpG-mediated pDC activation restored IFN-α secretion and increased CD8⁺ T cell-dependent tumor killing in vitro and in vivo, whereas pDC blockade (anti-BDCA-2) reversed these effects. Furthermore, we acknowledge that PRMT5 likely regulates SG resistance through additional tumor-intrinsic mechanisms, including epigenetic silencing of apoptosis-related genes and alterations in DNA damage repair pathways, and that the immune microenvironment-mediated mechanism described here represents one important dimension of a multifactorial resistance phenotype. Future studies integrating PRMT5 inhibition with immune checkpoint blockade or pDC activators will be necessary to fully characterize the extent and clinical relevance of this immune regulation axis. These findings supplement the mechanism by which PRMT5 regulates the tumor immune microenvironment. Moreover, through cell experiments and animal models, it was confirmed that knocking down PRMT5 in combination with SG treatment could significantly enhance the anti-tumor effects both in vitro and in vivo. Additionally, targeting Trop2 can reverse the suppressive state of CD8⁺ T cells [10]. The above evidence further clarifies the unique value of this combined strategy in reversing Trop2 ADC resistance, and provides a solid experimental basis for the clinical combined application of Trop2 ADC and pDC activators. It is worth noting that this study also discovered that PRMT5 undergoes nuclear-cytoplasmic translocation during SG resistance, and its enrichment in the cytoplasm is the key factor mediating pDC dysfunction. This change in subcellular localization provides a new direction for the subsequent development of drugs targeting the nuclear-to-cytoplasmic transport of PRMT5.

It is reported that the IRF3/I type interferon pathway can enhance the anti-tumor effect by promoting DC maturation and antigen-specific CD8⁺ T cell responses, including in TNBC, lung cancer and melanoma, and so on [42]. The study found that when PRMT5 was silenced, IRF3 was significantly upregulated, accompanied by the activation of pDCs. Given that the aforementioned study has confirmed that in SG-resistant TNBC tumor tissues, the expression of IRF3 is in a downregulated state, this study focuses on deeply exploring the specific mechanism by which PRMT5 regulates the expression of IRF3. Midline2 (MID2) is an E3 ubiquitin ligase of the TRIM family, which is encoded by the X chromosome. It is often highly expressed in breast cancer and can drive cell proliferation, mediate chemotherapy resistance, and indicate poor prognosis. It is a potential prognostic marker and therapeutic target [43]. In addition, Luo et al. showed that MORC4 could upregulate MID2 expression through STAT3, thereby promoting chemotherapy resistance in breast cancer [44]. This study confirmed that PRMT5 could recognize the RG motif of the MID2 protein and perform SDMA modification on it. The interaction between the methylated MID2 and IRF3 was enhanced, which in turn promoted the ubiquitination and degradation of IRF3, weakened the stability of IRF3, and thus significantly inhibited the recruitment and activation of pDCs. The elucidation of this mechanism links the epigenetic regulatory role of PRMT5 with the tumor immune microenvironment, revealing a new pathway mediated by PRMT5 for Trop2 ADC resistance. Meanwhile, this study also finds that MID2 is a key ubiquitin ligase for IRF3, and targeting MID2 may represent another potential strategy for reversing drug resistance.

This study still has certain limitations and needs to be further improved in subsequent research. Firstly, the clinical sample size of this study is relatively small: only 5 SG-sensitive (TNBC-S) and 5 SG-resistant (TNBC-R) patients were included. This limitation arises from the strict eligibility criteria — requiring histologically confirmed TNBC, documented clinical response to SG, and the availability of adequate tumor tissue specimens both before and after treatment — combined with the relatively recent approval and clinical introduction of SG. The limited sample size may affect the statistical power and generalizability of the clinical observations regarding pDC infiltration, IRF3 expression, and their correlation with SG sensitivity. While the mechanistic findings in this study are substantiated by multiple independent in vitro and in vivo experimental approaches, future multi-center prospective studies with larger patient cohorts are essential to validate the prognostic and predictive value of PRMT5 expression, pDC infiltration levels, and IRF3 protein status in TNBC patients receiving SG treatment. Secondly, this study did not delve deeply into the specific functional differences of PRMT5 in different localizations within the nucleus and cytoplasm. The molecular mechanism of its nuclear-cytoplasmic transport and the upstream regulatory pathways still need to be clarified. In the future, techniques such as nuclear-cytoplasmic separation and immunofluorescence co-localization can be employed, combined with high-throughput sequencing, to screen for key factors regulating the nuclear-cytoplasmic transport of PRMT5. Furthermore, the animal experiments used immune-deficient nude mouse models, which could not fully simulate the complex regulatory network of the human immune system. In particular, the roles of other immune cells, such as T cells and macrophages, in the combined treatment were not fully revealed. Subsequent studies should construct humanized mouse models or mouse models with a complete immune system to conduct research that is closer to clinical practice. Finally, this study only verified the role of the PRMT5-MID2-IRF3 axis in regulating the function of pDC. However, the mechanism of drug resistance in TNBC is highly heterogeneous, and whether PRMT5 participates in drug resistance regulation through other pathways still needs further exploration.

Taken together, PRMT5, as a key target connecting tumor cell resistance and immune microenvironment, is expected to become a breakthrough for the treatment of TNBC. Future research can focus on the following directions: First, developing highly specific PRMT5 inhibitors, especially targeted drugs for cytoplasmic PRMT5; second, optimize combination treatment regimens, exploring the synergistic effects of PRMT5 inhibitors with Trop2 ADC, immune checkpoint inhibitors, and pDC activators; third, discovering biomarkers based on PRMT5 expression levels for screening patients suitable for combined treatment. This study establishes a novel theoretical foundation for reversing Trop2 ADC resistance in TNBC and presents a valuable reference framework for investigating drug resistance mechanisms in other solid tumors, thereby holding substantial scientific and clinical importance.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Jing Kong, Minghui Li and Yidi Wang conducted the experiments and drafted the manuscript. Mengxuan Li collected and analyzed the data. Huimin Meng prepared figures, contributed the methodology, operated the software and edited the manuscript. Jiajun Ding, Jing Fan and Ting Wang designed and supervised the study. All authors reviewed the manuscript.

Funding

This study was supported by the Xijing Hospital Boost Project (XJZT25QN37 & XJZT26QN23).

Data availability

Data is available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Xijing Hospital Ethics Committee and complied with the Declaration of Helsinki. All patients signed informed consent. All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Xijing Hospital and were carried out strictly in compliance with pertinent guidelines.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.