MSRB3 promotes gastric cancer progression and immunotherapy resistance by stabilizing GPX4 to inhibit ferroptosis

Highlights

• •MSRB3 is overexpressed and predicts poor prognosis in gastric cancer.
• •MSRB3 suppresses ferroptosis by stabilizing GPX4 protein.
• •MSRB3 inhibits GPX4 ubiquitination and proteasomal degradation.
• •MSRB3 depletion increases CD8+ T-cell infiltration in tumors.
• •Targeting MSRB3 enhances response to anti-PD-1 immunotherapy.

Abstract

Methionine sulfoxide reductase B3 (MSRB3) possesses potent antioxidant activity and is highly expressed in various cancers, often correlating with poor prognosis. However, its biological effect and clinical importance in gastric cancer (GC) remain elusive. In this study, we first validated the expression and prognostic value of MSRB3 in GC through bioinformatics analyzes and clinical specimens. Subsequently, a series of functional experiments in vitro and in vivo were executed to explore the effect of MSRB3 in regulating GC cell proliferation, invasion, migration, ferroptosis, and the tumor immune microenvironment, as well as the underlying mechanisms. Finally, a subcutaneous xenograft model was established to assess the influence of MSRB3 on tumor growth and the efficacy of anti-PD-1 therapy. MSRB3 was markedly elevated in GC, and linked to poor patient prognosis. Functionally, MSRB3 facilitated the growth, migration, and invasion of GC cells. Indeed, MSRB3 stabilized GPX4 protein by inhibiting its ubiquitin-proteasomal-degradation, thereby suppressing lipid peroxidation and ferroptosis, which sustained GC cell survival and malignant progression. In vivo experiments confirmed that MSRB3 promoted tumor growth by inhibiting ferroptosis. Moreover, loss of MSRB3 increased intratumoral CD8⁺ T-cell accumulation and significantly enhanced the therapeutic response to PD-1 blockade. This synergy was reversed via the ferroptosis inhibitor Ferrostatin-1. MSRB3 functions as a suppressor of ferroptosis in GC by preserving GPX4 protein stability. MSRB3 drives tumor progression and confers immunotherapy resistance by stabilizing GPX4 to inhibit ferroptosis. Therefore, targeting MSRB3 may represent a potential strategy to improve immunotherapeutic efficacy in GC.

Graphical abstract

MSRB3 stabilizes GPX4 by inhibiting ubiquitination-mediated proteasomal degradation, thereby suppressing ferroptosis.

Introduction

Gastric cancer (GC) accounts for a substantial proportion of cancer-associated deaths worldwide, representing a major clinical challenge [1], Although therapeutic options for advanced GC have expanded over the past decade, overall treatment outcomes remain unsatisfactory for many patients [2]. Immune checkpoint blockade, particularly therapies targeting the PD-1/PD-L1 axis, has changed the therapeutic landscape for several malignancies [3]. However, its efficacy in GC is often limited and heterogeneous [4]. Only a fraction of patients with GC achieves durable clinical benefit, while a significant subset of patients exhibits primary or acquired resistance. Therefore, a deeper understanding of the underlying biological mechanisms is required to guide the development of improved therapeutic strategies.

Ferroptosis, a form of regulated cell death driven by iron-dependent lipid peroxidation, characterized by the accumulation of lipid reactive oxygen species (ROS) and membrane damage, has emerged as a critical vulnerability in cancer [5,6]. Central to the defense against ferroptosis is glutathione peroxidase 4 (GPX4), functioning as a key lipid repair enzyme that detoxifies lipid peroxides and maintains redox homeostasis [7]. Pharmacological inhibition or genetic depletion of GPX4 robustly induces ferroptosis, suppressing tumor growth [8]. Recent studies further reveal a crucial link between ferroptosis and antitumor immunity [9]. Induction of ferroptosis can enhance immunogenicity, reshape the tumor microenvironment, and improve responsiveness to immune checkpoint blockade [10]. For example, activated CD8⁺ T cells can inhibit system Xc− by secreting IFN-γ, thereby promoting ferroptosis in tumor cells [11]. Therefore, targeting ferroptosis regulators may provide an effective means of augmenting immunotherapeutic efficacy [12].

Methionine sulfoxide reductase B3 (MSRB3) catalyzes the reduction of methionine-R-sulfoxide residues, thereby contributing to the restoration of oxidatively modified proteins and maintenance of cellular redox homeostasis [13]. Previous studies have shown that MSRB3 is involved in the malignant progression of various cancers, including GC, and is associated with poor patient prognosis [[14], [15], [16], [17]]. Although its role in mitigating oxidative stress has been recognized [18], its specific function in cancer, particularly in regulating cell death pathways such as ferroptosis, remains poorly understood. Given the redox-sensitive nature of ferroptosis, we hypothesized that MSRB3 may serve as a novel regulatory node in this process, potentially through the stabilization of GPX4 or related protective proteins.

Our study identifies MSRB3 as a novel regulator of ferroptosis in GC. We demonstrate that MSRB3 promotes tumor progression by stabilizing GPX4 protein through inhibition of ubiquitination-mediated proteasomal degradation. Furthermore, ferroptosis induced by MSRB3 depletion promoted CD8⁺ T-cell infiltration and potentiated the antitumor efficacy of PD-1 blockade in GC. Our findings not only identify MSRB3 as a novel regulator of ferroptosis but also propose a promising therapeutic strategy to potentiate immunotherapy for GC.

Materials and methods

Bioinformatics analysis

Gene expression datasets GSE122401, GSE54129, and GSE184336 were obtained from the Gene Expression Omnibus (GEO) database. The transcriptomic dataset TCGA-STAD was sourced from The Cancer Genome Atlas (TCGA) program. Bioinformatics analyzes were performed using the following platforms: ASSISTANT for Clinical Bioinformatics (www.aclbi.com), ZeroCoding (www.zc.bioinfosci.cn), GEPIA (gepia.cancer-pku.cn) and TIMER 3.0 (compbio.cn/timer3).

Specimens and ethics

Primary GC tissues and paired adjoining healthy tissues were supplied by 30 consecutive patients with histologically confirmed GC at Jiangmen Central Hospital. This study was approved by the Institutional Review Board of Jiangmen Central Hospital (Approval No 202,270), and written informed consent was obtained from all participants in accordance with the Declaration of Helsinki.

Cell culture

Human GC cells (SNU1, HGC27, MKN28 and AGS), normal human gastric mucosal epithelial cell line (GES1) and mouse GC cell line (MFC) were provided by the American Type Culture Collection (ATCC). SNU1, MKN28, MFC, and GES1 were cultured in RPMI-1640 medium. HGC27 and AGS were maintained in Dulbecco's Modified Eagle Medium (DMEM). All were supplemented with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin and maintained under 5% CO₂ at 37 °C.

Antibodies and reagents

The antibodies to MSRB3 (14,251-1-AP), CD8a (66,868-1-Ig), CD8a (29,896-1-AP) and β-actin (60,004-1-Ig) were offered by Proteintech (Rosemont, IL, USA). The antibody to GPX4 (#59,735) was obtained from Cell Signaling Technology (Danvers, MA, USA). Anti-PD-1 antibody (RMP1–14) was purchased from BioXCell (West Lebanon, NH, USA). Ferrostatin-1 (HY-100,579), MG132 (HY-13,259), Bafilomycin A1 (HY-100,558) and chloroquine (HY-17589A) were purchased from MCE (Monmouth Junction, NJ, USA).

Construction of shRNA plasmids and lentivirus production

The MSRB3 overexpression plasmid was obtained from the Public Protein/Plasmid Library. To knock down MSRB3 expression, short hairpin RNA (shRNA) sequences targeting MSRB3 were cloned into the lentiviral vector pLKO.1 (#8453, Addgene). For lentivirus packaging, the constructed shRNA plasmids were co-transfected with psPAX2 and pMD2.G (P0261 and P0262, MiaoLing Biology) into HEK293T cells via Turbofect transfection reagent (R0531, Thermo Fisher Scientific, Waltham, MA, USA). The viral supernatant was collected and utilized for infecting GC cells under the polybrene. Then, stable cell lines were formulated by selecting with the appropriate antibiotics. Resistant monoclonal cell lines were expanded and used for subsequent functional experiments. The targeting sequences of the shRNAs are listed below: shMSRB3#1: GTGTTGTTTGTGGAACTCCAT; shMSRB3#2: CATTCCACGATGTGATCAATT.

Cell viability assays

GC cells were seeded at 3.0 × 10³ cells per well in 96-well plates (triplicate wells per condition) and incubated overnight. Cells were treated with 1 μM Ferrostatin-1 for 24 h prior to functional assays. Cell viability was assessed using the CCK-8 assay. Briefly, CCK8 reagent was added to each well at a final concentration of 10%, and cells were incubated at 37 °C for 1 h. Absorbance at 450 nm was measured using a Varioskan LUX multimode microplate reader (Thermo Fisher Scientific, Waltham, MA, USA) with detection time points at 0, 24, 48, and 72 h.

Colony formation assays

GC cells were seeded at 1.0 × 10³ cells each well within 6-well plates (triplicate wells per condition) and cultivated for 2 weeks. Colonies were fixed with methanol for 30 min, stained with 0.1% crystal violet (Beyotime Biotechnology, Shanghai, China) for 10 min, and quantified using ImageJ software.

Ethynyl deoxyuridine (EdU) assay

Cell proliferation was assessed via a Cell-Light™ EdU Apollo staining kit (RiboBio, Guangzhou, China). EdU incorporation was visualized by confocal microscopy (Zeiss, Oberkochen, Germany) and quantified as a percentage of total cells.

Wound healing assay

The cell monolayers were scratched with a sterile pipette tip and cultured in serum-free medium after removal of cellular debris. Image capture was executed at the indicated time points. Wound closure was assessed via gaging the remaining gap area.

Transwell invasion assays

Cell invasion was evaluated via Matrigel-coated Transwell chambers (8-μm pore size). Cells (2.0 × 10⁵ per well) were suspended in serum-free medium and seeded into the upper chamber. Medium supplemented with 10% FBS was added to the lower chamber. Following cultivation, the cells traversed the membrane, were immobilized, stained with crystal violet, and counted in five randomly selected microscopic fields.

Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted using TRIzol reagent (Invitrogen, 15,596,026) and reverse-transcribed into cDNA using a reverse transcription kit (Servicebio, G3329–50). qRT-PCR was performed using SYBR Green Master Mix (Servicebio, G3320–01) on a real-time PCR system. Transcript abundance was normalized to GAPDH and analyzed using the 2−ΔΔCt method. The primers sequences were as follows: GPX4-F: CAGTGAGGCAAGACCGAAGT, GPX4-R: GGGGCAGGTCCTTCTCTATC, GAPDH-F: AGCCACATCGCTCAGACAC, GAPDH-R: GCCCAATACGACCAAATCC.

Western blot analysis

Whole-cell lysates were fabricated within RIPA buffer comprising protease inhibitors. Protein samples were resolved via SDS-PAGE, delivered onto PVDF membranes, and probed with antibodies against MSRB3, GPX4, and β-actin, before HRP-conjugated secondary antibodies. Immunoreactive bands were observed through improved chemiluminescence.

Immunofluorescence

GC cells were immobilized with 4% paraformaldehyde, permeabilized with Triton X-100, and then sealed with bovine serum albumin, before nurtured with primary and fluorophore-conjugated secondary antibodies. Nuclei were stained with DAPI, and fluorescence signals were documented microscopically.

Immunohistochemistry (IHC)

Formalin-fixed, paraffin-embedded tissue sections were deparaffinized and rehydrated, followed by citrate-based antigen retrieval (pH 6.0). After blockade of endogenous peroxidase viability and nonspecific protein binding, the sections were cultivated with anti-MSRB3 or anti-GPX4 antibodies overnight at 4 °C. After washing, the sections were incubated with HRP-conjugated secondary antibodies. Immunoreactivity was visualized using DAB and counterstained with hematoxylin. The sections were subsequently rinsed, dehydrated through a graded ethanol series, cleared in xylene, and mounted for microscopic examination.

Reactive oxygen species (ROS) detection

Intracellular ROS levels were measured using the DCFH-DA (Beyotime, S0033, Shanghai, China) according to the manufacturer's protocol. Cells were incubated with DCFH-DA in serum-free medium at 37 °C for 30 min, and fluorescence intensity was quantified by flow cytometry.

Lipid peroxidation assay

Intracellular lipid peroxidation was assessed by determining malondialdehyde (MDA) using a Lipid Peroxidation MDA Assay Kit (Beyotime, S0131, Shanghai, China).

Glutathione and intracellular Fe²⁺ measurement

Intracellular ferrous iron (Fe²⁺) levels were visualized using FerroOrange probe (Dojindo, Japan). Following probe loading, fluorescence was examined by confocal microscopy (Zeiss). Cellular glutathione (GSH) levels were quantified via a GSH Assay Kit (Nanjing Jiancheng Bioengineering Institute, A006–2–1, Nanjing, China).

Transmission electron microscopy (TEM)

Cells were immobilized with electron microscopy fixative (G1102, Servicebio, Wuhan, China) at room temperature (RT) for 5 min in darkness, gently scraped off and transferred to a centrifuge tube. After low-speed centrifugation for 3 min, the cell pellet was suspended in fresh EM fixative again and further fixed at RT for 30 min in darkness, before stored at 4 °C. The samples were subsequently sent to Servicebio (Wuhan, China) for sectioning and imaging. Transmission electron microscopy images were acquired using an HT7800 microscope (Hitachi, Tokyo, Japan).

Cycloheximide chase assay

For cycloheximide (CHX) chase assays, cells with MSRB3 overexpression or knockdown and the corresponding control cells were treated with CHX (50 μg/mL). Cells were harvested at 0, 6, 18, and 24 h after CHX treatment, and GPX4 protein levels were determined by western blotting. GPX4 abundance was normalized to β-actin and expressed relative to the level at 0 h.

Ubiquitination assay

Cells were exposed to MG132 (10 μM) for 6 h before collection. Cell lysates containing protease and deubiquitinase inhibitors underwent immunoprecipitation with the indicated antibody and Protein A/G beads. Ubiquitinated proteins were subsequently eluted and detected by western blot analysis.

Animal experiments

Female BALB/c nude and BALB/c mice (5–6 weeks old) were purchased from Guangdong Yaokang Co., Ltd. All procedures were conducted in accordance with the Animal Ethics Committee of Jinan University (Approval No K-20,240,603-06). A subcutaneous xenograft mouse model was established by injecting 5 × 10⁶ GC cells into the flank of each mouse. When the tumor volume reached approximately 80 mm³, mice were randomly assigned to groups for subsequent treatments. In the in vivo experiments, Fer-1 and RMP1–14 were administered intraperitoneally at a dosage of 10 mg/kg every third day. Tumor size was measured every 2 days, and its volume was calculated using the formula: tumor length × tumor width² / 2. Tumors were excised at the experimental endpoint for subsequent analyzes.

Statistical analysis

All assays were executed in three replicates unless otherwise indicated. Data are denoted as mean ± standard deviation (SD). Statistical analyzes were executed via GraphPad Prism 8 (GraphPad Software, La Jolla, CA, USA). Differences between two groups were construed via Student’s t-test, while multiple groups were contrasted via one-way ANOVA. A two-sided P value < 0.05 held statistical significance.

Results

MSRB3 predicts poor prognosis in gastric cancer

To investigate the clinical relevance of MSRB3 in gastric cancer, three independent GEO datasets (GSE122401, GSE54129 and GSE184336) were analyzed, revealing that MSRB3 expression was remarkably upregulated in GC tissues relative to adjacent normal tissues (Fig. 1A–C). Consistently, analysis of the TCGA-STAD dataset exhibited that elevated MSRB3 expression was positively related to deeper tumor invasion (T stage) and more advanced pathological stage (Fig. 1D,E). More importantly, Kaplan–Meier analysis of the TCGA-STAD dataset using GEPIA demonstrated that GC patients showing a high MSRB3 expression possessed a significantly shorter overall survival (OS) than those with low MSRB3 expression (Fig. 1F). To validate the bioinformatics results, we first examined MSRB3 expression in surgical specimens obtained from 30 GC patients. Consistent with the GEO data, MSRB3 protein was significantly upregulated in tumor tissues relative to adjacent normal tissues (Fig. 1G). We further validated this finding at the cellular level, showing that both mRNA and protein levels of MSRB3 were markedly higher in GC cell lines than in normal gastric epithelial cells (Fig. 1H,I). Collectively, these findings suggest that MSRB3 can be considered a key factor associated with poor prognosis in GC patients.

Fig. 1.

MSRB3 predicts poor prognosis in gastric cancer. (A-C) MSRB3 expression was analyzed in gastric cancer versus normal tissues using the GSE122401, GSE54129, and GSE184336 datasets. (D-E) Relationship between MSRB3 expression and T stage or pathological stages. (F) Kaplan–Meier analysis of overall survival (OS) according to MSRB3 expression level. (G) Representative images of IHC staining for MSRB3 in GC tissues and their adjacent normal tissues, Scale bars, 200 μm and 50 μm. Quantitative analysis was shown in the graphs. (H, I) Wester blotting and qRT-PCR analysis of MSRB3 expression in GC cell lines (MKN28, SNU1, AGS and HGC27) and normal gastric epithelial cells (GSE1). *P < 0.05, **P < 0.01, ***P < 0.001.

MSRB3 promotes gastric cancer cells proliferation, migration, and invasion

To define the functional role of MSRB3 in GC, MKN28 and AGS cells, which exhibited intermediate endogenous MSRB3 expression among the examined GC cell lines, were selected to provide an appropriate experimental window for MSRB3 gain- and loss-of-function analyzes. We first performed gain-of-function experiments in these two cell lines (Fig. 2A). MSRB3 overexpression markedly improved cell viability, as determined by CCK-8 assays (Fig. 2B), and increased clonogenic growth (Fig. 2C). Consistently, EdU incorporation assays revealed a higher proportion of proliferating cells upon MSRB3 overexpression (Fig. 2D,E), indicating a robust pro-proliferative effect. We next examined whether MSRB3 influences the migratory and invasive capacities of GC cells. Wound healing assays demonstrated that MSRB3 overexpression significantly accelerated wound closure at 24 h (Fig. 2F). Consistently, Transwell invasion assays showed a marked increase in the number of invaded cells (Fig. 2G), supporting a role for MSRB3 in promoting the invasive phenotype of GC cells.

Fig. 2.

MSRB3 promotes gastric cancer cell proliferation, migration, and invasion. (A) MSRB3 overexpression in MKN28 and AGS cells were validated by western blotting. (B) Cell viability assessed by CCK-8 assays following MSRB3 overexpression. (C) Representative images and quantitative analysis of colony formation assays. (D, E) EdU incorporation assays and quantification of EdU-positive cells. Scale bar, 50 μm. (F) Wound healing assays were conducted to assess cell migration ability in MKN28 and AGS cells at 24 h. Scale bar, 200 μm. (G) Transwell invasion assays with quantification of invaded cells. Scale bar, 100 μm. Data are presented as mean ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

MSRB3 regulates ferroptosis among GC cells

Given the antioxidant function of MSRB3 and the critical role of lipid peroxidation in ferroptosis, we investigated whether MSRB3 regulates ferroptotic cell death in GC. First, we evaluated ferroptosis-related markers in GC cells overexpressing MSRB3. Compared with control cells, MSRB3-overexpressing cells showed significantly decreased levels of ROS, Fe²⁺, and MDA, and a marked increase in GSH (Fig. 3A–D). The expression of GPX4, a key ferroptosis suppressor, was also notably upregulated in these cells (Fig. 3E).

Fig. 3.

MSRB3 suppresses ferroptosis in gastric cancer cells. (A–D) Intracellular ROS (A), Fe²⁺ (blue, DAPI-stained nuclei; yellow, Fe²⁺) (B), MDA (C), and GSH levels (D) following MSRB3 overexpression. Scale bar, 50 μm and 20 μm. (E) Western blotting analysis of MSRB3 and GPX4 expression, with quantification of GPX4 protein levels upon MSRB3 overexpression. Data are presented as mean ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

To assess the contribution of ferroptosis to MSRB3-driven GC progression, rescue experiments were performed using Ferrostatin-1 (Fer-1) in MSRB3-knockdown cells (Fig. 4A). Fer-1 largely reversed the inhibition of cell viability, proliferation, migration, and invasion induced by MSRB3 knockdown (Fig. 4B–F). Furthermore, Fer-1 counteracted the ferroptotic changes induced by knockdown of MSRB3, as evidenced by reduced ROS, Fe²⁺, and MDA accumulation together with recovery of intracellular GSH levels (Fig. 4G–J). Consistently, transmission electron microscopy revealed characteristic ferroptotic mitochondrial damage, including shrinkage, loss of cristae, and membrane rupture, in MSRB3‑knockdown cells; these morphological changes were attenuated by Fer-1 treatment (Fig. 4K). Finally, GPX4 protein levels decreased upon MSRB3 knockdown and were partially restored upon Fer-1 addition (Fig. 4L).

Fig. 4.

MSRB3 deficiency induces ferroptosis in gastric cancer cells. (A) Knockdown efficiency of MSRB3 in MKN28 and AGS cells was confirmed by western blotting. (B) Cell viability following MSRB3 depletion with or without Fer-1 (1 μM). (C) Colony formation assays showing that treatment with Fer-1 (1 μM) restored clonogenic capacity in MSRB3-deficient cells. Quantification of colony formation in MSRB3 knockdown or control group. (D) Representative EdU incorporation images and quantification under indicated conditions. Scale bar, 50 μm. (E) Wound healing assays demonstrating that Fer-1 (1 μM) treatment rescued impaired migratory ability caused by MSRB3 knockdown. Quantification of migration distance under indicated conditions. Scale bar, 200 μm. (F) Transwell invasion assays showing that Fer-1 (1 μM) reversed migration inhibition in MSRB3-deficient cells. Scale bar, 100 μm. (G) Intracellular ROS levels were measured by DCFH-DA staining. Scale bar, 50 μm. (H) Lipid peroxidation levels were measured by MDA assay under indicated conditions. (I) Intracellular GSH levels were quantified under indicated conditions. (J) Intracellular Fe²⁺ levels were measured using iron probes in MKN28 and AGS cells (blue, DAPI-stained nuclei; yellow, Fe²⁺). Scale bar, 20 μm. (K) Transmission electron microscopy revealed characteristic ferroptotic mitochondrial (red arrows), including mitochondrial shrinkage, increased membrane density, and loss of cristae. Scale bar, 500 nm. (L) GPX4 protein expression analyzed by western blotting with quantification under indicated conditions. Data are presented as mean ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Together, these data demonstrate that MSRB3 functions as a potent ferroptosis suppressor, thereby sustaining GC cell survival and malignant progression.

MSRB3 stabilizes GPX4 by inhibiting ubiquitination-mediated proteasomal degradation

Given the established regulatory role of MSRB3 on GPX4 and the central position of GPX4 in ferroptosis, we next sought to investigate how MSRB3 regulates GPX4 expression. We first considered its regulation at the transcriptional level and found that MSRB3 had no significant effect on GPX4 mRNA expression (Fig. 5A), suggesting a post-transcriptional level mechanism. To determine whether MSRB3 impacts GPX4 protein stability, we performed cycloheximide (CHX) chase assays. MSRB3 overexpression significantly prolonged the half-life of GPX4 protein, whereas MSRB3 depletion accelerated GPX4 degradation (Fig. 5B, C), indicating that MSRB3 stabilizes GPX4 at the protein level. To clarify how GPX4 is destabilized following MSRB3 depletion, cells were exposed to inhibitors targeting different proteolytic pathways. Blockade of the proteosome with MG132, but not the autophagy inhibitor chloroquine (CQ) or the lysosomal inhibitor bafilomycin A1 (Baf-A1), restored GPX4 abundance in MSRB3-deficient cells (Fig. 5D). Correspondingly, MSRB3 depletion promoted GPX4 ubiquitination, while MSRB3 overexpression diminished ubiquitin accumulation on GPX4 (Fig. 5E).

Fig. 5.

MSRB3 stabilizes GPX4 by inhibiting ubiquitination-mediated proteasomal degradation. (A) GPX4 mRNA levels were assessed by qRT-PCR following MSRB3 overexpression or knockdown in GC cells. (B, C) CHX (50 μg/ml) chase assays were performed to evaluate GPX4 protein stability over time in overexpression MSRB3 or knockdown group at 0, 6, 18, and 24 h (B). Quantification of GPX4 protein levels (C). (D) MKN28 and AGS cells were treated with the proteasome inhibitor MG132 (10 μM), autophagy inhibitor CQ (20 μM), or lysosomal inhibitor Baf-A1 (200 nM), and GPX4 protein levels were determined by western blotting. Quantification of GPX4 protein levels treated by MG132, CQ, or Baf-A1. (E) Ubiquitination assays showing GPX4 ubiquitination levels upon MSRB3 modulation, with quantification relative to total GPX4. Data are presented as mean ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Briefly, our results recognize MSRB3 as a critical modulator of GPX4 protein stability, functioning by inhibiting ubiquitination-mediated proteasomal degradation.

MSRB3 regulates ferroptosis and tumor progression in vivo

To evaluate the in vivo role of MSRB3, subcutaneous xenograft models were established using GC cells with stable MSRB3 overexpression or knockdown. MSRB3 overexpression significantly accelerated tumor growth, as evidenced by increased tumor volume and weight (Fig. 6A–C). Consistent with our in vitro findings, MSRB3 overexpression reduced lipid peroxidation and enhanced antioxidant capacity in tumor tissues, as indicated by increased GSH levels and decreased MDA levels (Fig. 6D–E). In contrast, MSRB3 knockdown markedly suppressed tumor progression (Fig. 6G–I). MSRB3 depletion resulted in reduced GSH levels and elevated lipid peroxidation (Fig. 6J–K). Importantly, treatment with Fer-1 partially rescued tumor growth inhibition induced by MSRB3 knockdown, supporting a ferroptosis-dependent mechanism in vivo. Immunohistochemical analysis further confirmed that MSRB3 expression positively correlated with GPX4 levels in tumor tissues (Fig. 6F,L), reinforcing the MSRB3–GPX4 regulatory axis. Together, these in vivo data demonstrate that MSRB3 promotes GC progression by suppressing ferroptosis through stabilization of GPX4.

Fig. 6.

MSRB3 regulates ferroptosis and tumor progression in vivo. (A–F) MSRB3 overexpression inhibits ferroptosis and enhances tumor growth in vivo. (A) Representative images of xenograft tumors derived from control (Vector) and MSRB3-overexpressing GC cells. (B) Tumor growth curves. (C) Final tumor weights at endpoint. (D, E) Relative GSH levels (D) and MDA levels (E) in xenografts. (F) Representative IHC staining of GPX4 and Ki-67 in tumor sections, with quantification of positive staining. Scale bar, 50 μm. (G–L) MSRB3 knockdown suppresses tumor growth in a ferroptosis-dependent manner. (G) Representative images of xenografts from Src+DMSO, shMSRB3+DMSO, and shMSRB3+Fer-1 groups. (H) Tumor growth curves. (I) Tumor weights at endpoint. (J, K) Relative GSH levels (J) and MDA levels (K) in indicated groups under indicated conditions. (L) Representative IHC staining and quantification of GPX4 and Ki-67 in tumor sections. Scale bar, 50 μm. Data are presented as mean ± SD (n = 5 mice per group). *P < 0.05, **P < 0.01, ***P < 0.001.

MSRB3 deficiency enhances anti-PD-1 efficacy in gastric cancer by inducing ferroptosis

Given the emerging interplay between ferroptosis and antitumor immunity, we explored whether MSRB3 influences the immune landscape of GC [11,19,20]. Gene set enrichment analysis revealed GC samples with high MSRB3 expression were enriched in immune suppressive–related signatures, particularly “negative regulation of immune effector process” (Fig. 7A). This demonstrates that the high MSRB3 expression might be linked to a locally immunosuppressive state in tumors. In contrast, GC samples with low MSRB3 expression specifically showed enrichment in “CD8-positive αβ T cell homeostasis” (Fig. 7B). Consistently, analysis using the TIMER3.0 database exhibited a negative relationship between MSRB3 expression levels and CD8⁺ T cell infiltration by two independent algorithms (Fig. 7C,D). Similar observations were obtained in clinical GC specimens, where markedly raised CD8⁺ T-cell infiltration among tumors with low MSRB3 expression relative to those showing a high MSRB3 expression (Fig. 7E,F). Together, these findings link reduced MSRB3 expression to a more T-cell-inflamed tumor microenvironment.

Fig. 7.

MSRB3 expression is negatively associated with CD8⁺ T-cell infiltration in gastric cancer. (A, B) Gene Ontology Enrichment Analysis (GO) showing enrichment of immune suppressive pathways in the MSRB3-high group, whereas the MSRB3-low group was enriched in the “CD8-positive αβ T-cell homeostasis” pathway. (C, D) TIMER3.0 analysis showing a negative correlation between MSRB3 expression and CD8⁺ T-cell infiltration in TCGA-STAD samples. (E) Representative immunofluorescence staining of MSRB3 (green), CD8 (red), and DAPI (blue) in GC tissues with high or low MSRB3 expression. Scale bar, 25 μm. (F) Quantification of infiltrating CD8⁺ T cells in GC tissues with high or low MSRB3 expression. Data are presented as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001.

To determine whether MSRB3 deficiency-induced ferroptosis could enhance the efficacy of immunotherapy against GC, we established a subcutaneous xenograft tumor model using shMSRB3-MFC cell lines (Fig. 8A). The results showed that both MSRB3 knockdown alone and anti-PD-1 antibody (RMP1–14) monotherapy significantly suppressed tumor growth compared with the control group, as evidenced by reduced tumor volume, tumor weight, and Ki-67 positive areas. Moreover, combined MSRB3 knockdown and anti-PD-1 treatment exerted a markedly stronger antitumor effect than either monotherapy, indicating a synergistic therapeutic efficacy. Notably, this therapeutic advantage was largely abolished following administration of the ferroptosis inhibitor Fer-1, which restored tumor growth to levels comparable to those observed with single-agent treatment (Fig. 8B–F).

Fig. 8.

MSRB3 deficiency enhances anti-PD-1 efficacy in a ferroptosis-dependent manner and augments CD8⁺ T cell infiltration in vivo. (A) Knockdown efficiency of MSRB3 in MFC cells was confirmed by western blotting. (B–D) Representative tumor images (B), tumor growth curves (C), and tumor weights (D) of xenograft models treated with anti–PD-1 antibody and/or Fer-1. MSRB3 knockdown synergized with PD-1 blockade to suppress tumor growth, which was reversed by Fer-1 treatment. (E, F) Representative Ki-67 staining and quantification showing markedly reduced tumor proliferation in the combination treatment group. Scale bar, 50 μm. (G, H) Representative immunofluorescence images and quantification of CD8⁺ T-cell infiltration in xenograft tumors. Scale bar, 50 μm. Data are presented as mean ± SD (n = 5 mice per group). *P < 0.05, **P < 0.01, ***P < 0.001.

Assessment of intratumoral immune infiltration revealed a parallel pattern. Anti-PD-1 treatment raised CD8⁺ T-cell accumulation relative to controls, whereas the combination regimen elicited the maximal level of CD8⁺ T-cell infiltration among all groups. This increase was markedly attenuated by Fer-1, diminishing CD8⁺ T-cell infiltration to a level comparable to that in anti-PD-1 monotherapy group (Fig. 8G,H). Collectively, these results indicate that MSRB3 deficiency enhances ferroptosis-dependent antitumor immune responses and improves the responsiveness of GC to PD-1 blockade.

Next, to explore the immunological mechanism by which MSRB3 knockdown enhances anti-PD-1 efficacy, we assessed CD8⁺ T cell infiltration within the xenograft tumors. IF staining revealed that anti-PD-1 monotherapy increased CD8⁺ T-cell infiltration relative to controls. Importantly, the combination treatment group exhibited substantially greater CD8⁺ T-cell infiltration than either monotherapy group, whereas Fer-1 treatment significantly attenuated this enhancement, reducing CD8⁺ T-cell infiltration to levels comparable to those visualized in the anti-PD-1 monotherapy group (Fig. 8G,H). Collectively, these data demonstrate that MSRB3 knockdown enhances ferroptosis-mediated antitumor immunity, promotes CD8⁺ T-cell infiltration, while sensitizing GC to anti-PD-1 immunotherapy.

Discussion

GC progression is tightly coupled to adaptive redox homeostasis, making tumor cells evade oxidative damage. In this study, we identify MSRB3 as a critical determinant of ferroptosis resistance in GC. Mechanistically, MSRB3 stabilizes GPX4 by preventing its ubiquitin-dependent proteasomal degradation, thereby suppressing ferroptosis and promoting malignant progression. MSRB3, a methionine-R-sulfoxide reductase, safeguards proteome integrity under oxidative stress [21]. Beyond this antioxidant function, accumulating evidence implicates MSRB3 in oncogenic progression across multiple cancer types [14,22,23]. Consistent with this, our data confirm its upregulation and prognostic significance in GC. Functionally, MSRB3 promotes GC cell proliferation, migration, and invasion. Importantly, rescue experiments with Fer-1 demonstrated that these oncogenic phenotypes are largely mediated through ferroptosis inhibition.

Mechanistically, we delineate that MSRB3 regulates ferroptosis suppressor GPX4 at the post-translational level. MSRB3 depletion accelerated GPX4 degradation without affecting GPX4 mRNA expression, and this effect was rescued by MG132 but not Baf-A1, demonstrating that MSRB3 maintains GPX4 stability through inhibition of ubiquitin-mediated proteasomal degradation. These results identify MSRB3 as a key modulator of GPX4 turnover and highlight its role in ferroptosis regulation. Our findings further suggest that redox adaptation and ubiquitin-mediated GPX4 turnover are functionally interconnected during ferroptosis resistance [[24], [25], [26], [27]].

A significant translational implication of this study lies in the connection between ferroptosis and antitumor immunity. Although immune checkpoint blockade has improved the treatment landscape of GC, the overall response rate to anti–PD-1/PD-L1 treatment is still limited due to profound immunosuppression within the tumor microenvironment. Increasing evidence indicates that ferroptotic tumor cells can promote antitumor immunity by enhancing CD8⁺ T-cell recruitment and activation. Through the release of DAMPs, such as HMGB1 and ATP, ferroptotic tumor cells may promote dendritic cell activation and amplify downstream cytotoxic T-cell-mediated immunity [28,29]. Consistently, we observed a significant negative relationship between MSRB3 expression and CD8⁺ T-cell infiltration in TIMER3.0 analyzes and clinical GC tissues. More importantly, MSRB3 knockdown markedly enhanced the antitumor efficacy of PD-1 blockade, accompanied by increased intratumoral CD8⁺ T-cell infiltration, whereas these effects were largely abolished by Fer-1, confirming a ferroptosis-dependent mechanism. Mechanistically, ferroptosis induced by MSRB3 depletion likely remodels the tumor microenvironment into a more pro-inflammatory and immune-permissive state, thereby alleviating local immunosuppression and facilitating the recruitment and infiltration of activated CD8⁺ T cells into tumor tissues [20]. PD-1 blockade may further sustain the effector function of these infiltrating T cells by preventing T-cell exhaustion, ultimately maximizing their cytotoxic antitumor activity [11,30,31].

Several limitations should be acknowledged. First, although we demonstrated that MSRB3 is upregulated in GC and associated with poor prognosis, we did not stratify patients by TCGA molecular subtypes (EBV-positive, MSI, genomically stable, and chromosomally unstable). Given the known heterogeneity of GC, MSRB3 expression and its prognostic impact may vary across subtypes, warranting future validation in larger, subtype-stratified cohorts. Second, while our data show that MSRB3 enhances GPX4 protein stability, the underlying biochemical mechanism remains unclear. As a methionine sulfoxide reductase, MSRB3 could stabilize GPX4 by reversing oxidative damage on methionine residues, or alternatively, by modulating components of the ubiquitin-proteasome system. Although TRIM21, TRIM25, TRIM46, and NEDD4L have been reported as E3 ligases targeting GPX4 [32,33], the specific ligase(s) or deubiquitinase(s) mediating MSRB3-regulated GPX4 turnover have not been identified. Third, although our in vitro wound-healing and Transwell assays support a role for MSRB3 in promoting GC cell migration and invasion, this phenotype was not validated using orthotopic or experimental metastasis models in vivo. Therefore, the contribution of MSRB3 to metastatic dissemination in vivo remains to be established in future studies. Furthermore, whether MSRB3 directly interacts with GPX4 or acts through adaptor proteins requires experimental confirmation. Finally, future studies with expanded clinical cohorts are essential to establish MSRB3 as a prognostic biomarker and therapeutic target in GC.

In summary, we identify a previously unrecognized MSRB3–GPX4 regulatory axis in which MSRB3 stabilizes GPX4 by inhibiting its ubiquitin-mediated degradation, thereby suppressing ferroptosis and promoting GC progression. Importantly, targeting MSRB3 enhances ferroptosis-dependent CD8⁺ T-cell infiltration and sensitizes GC to PD-1 blockade, highlighting ferroptosis regulation as a critical determinant of antitumor immunity and immunotherapeutic responsiveness in GC.

Funding

This work was supported by National Natural Science Foundation of China (82173236) and the Medical Science and Technology Research Fund Project of Guangdong (B2024073).

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Ethics declarations

This study was performed in line with the principles of the Declaration of Helsinki. Approval for animal experiment was granted by the Animal Ethics Committee of Jinan University (approval no K-20,240,603-06). Ethical approval for the collection of clinical samples was obtained from the Ethics Committee of Jiangmen Central Hospital (approval no 202,270).

CRediT authorship contribution statement

Jianqin Lai: Writing – review & editing, Writing – original draft, Methodology, Investigation, Conceptualization. Taiwei Mo: Writing – review & editing, Methodology, Investigation. Jinxuan Su: Writing – review & editing, Writing – original draft, Investigation. Zhijing Zhang: Writing – review & editing, Formal analysis, Data curation. Xiaofeng Zhou: Writing – review & editing, Formal analysis, Data curation. Chengxing Wang: Resources. Qiuming Zou: Writing – review & editing, Formal analysis, Data curation. Sijia Li: Writing – review & editing, Formal analysis, Data curation. Xiaosong Zhuang: Writing – review & editing, Formal analysis, Data curation. Fangyuan Chen: Writing – review & editing, Methodology, Conceptualization. Qing Huang: Supervision, Conceptualization. Jie Cao: Writing – review & editing, Supervision, Methodology, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Our data can be obtained from the corresponding author if requested reasonably.