GPX4 knockdown suppresses M2 macrophage polarization in gastric cancer by modulating kynurenine metabolism

Abstract

Background: Glutathione peroxidase 4 (GPX4), an important factor regulating redox homeostasis, plays an important role in tumor microenvironment and progression. However, the role of GPX4 in gastric cancer (GC) is unclear. Methods: Spectral flow cytometry and multiplex immunohistochemistry were employed to assess the correlation between GPX4 expression and immune cell infiltration. Metabolomics analysis of conditioned media from GPX4 knockdown NUGC3 cells identified metabolic alterations. Additionally, both in vitro and in vivo functional studies were conducted to elucidate the mechanistic role of GPX4 in regulating the tumor microenvironment and progression. Results: Knockdown of GPX4 in GC cells inhibited tumor growth, enhanced CD8⁺ T cell infiltration, and suppressed the polarization of tumor-associated macrophages (TAMs) toward the pro-tumor M2 phenotype. Multiplex immunohistochemistry revealed a positive correlation between GPX4 expression and M2 macrophage infiltration in clinical samples from patients with GC. Metabolomics revealed that GPX4 knockdown regulate kynurenine metabolism pathway. Furthermore, mechanistic studies reveal that GPX4 silencing elevates lipid peroxidation, triggering the conversion of KYNU ubiquitin chain modifications from K48 to K63. Such ubiquitination remodeling stabilizes KYNU expression (a key kynurenine-metabolizing enzyme), reduces kynurenine accumulation, and ultimately reprograms TAM polarization to enhance antitumor immunity. We also identified that the K96 and K163 sites are important for KYNU's modification by K48 and K63 ubiquitin chains. Conclusion: Our study not only affirm the role of GPx4 in GC progression but also highlight it as a promising target for reshaping the immune microenvironment.

Keywords: GPX4; KYNU; TAM polarization; gastric cancer; kynurenine metabolism; ubiquitin.

Figure 1

Impacts of GPX4 knockdown on tumor growth, immune cell infiltration, and patient prognosis in gastric cancer. (A) Differential gene expression analysis was performed to assess ferroptosis-related pathways by transcriptomic sequencing of 60 pairs of gastric cancer (GC) and adjacent noncancerous tissue specimens from Zhejiang Cancer Hospital. (B) Diagram of the experimental design to assess the effect of GPX4 knockdown on immune cell infiltration in a murine GC cell line (MFC). (C-D) Tumor growth (tumor images and weights) in mice was monitored following GPX4 knockdown in MFC cells compared to the control group. (E) t-Distributed Stochastic Neighbor Embedding (t-SNE) plots generated from spectral flow cytometry data to visualize immune cell populations. (F-G) Impacts of GPX4 knockdown on tumor-associated macrophage (TAM) populations, M1 and M2 macrophages, assessed by spectral flow cytometry. (H) A tissue microarray was constructed with tumor samples from 179 GC patients, and multiplex immunohistochemistry was performed to examine the relationship between GPX4 expression and M2/M1 macrophage infiltration ratio within the tumor tissue. (I) Representative images of multiplex immunohistochemical staining illustrating the differential infiltration of M1 and M2 macrophages in tumor tissue samples with varying GPX4 expression levels. Bar = 50 μm (J) Kaplan-Meier survival analysis conducted to evaluate the prognostic significance of GPX4 expression and macrophage infiltration.

Figure 2

GPX4 knockdown modulates macrophage polarization and suppresses tumor growth in gastric cancer. (A) Schematic representation of macrophage depletion using Clodronate Liposomes (CL) in mice to evaluate the role of macrophages in GPX4 knockdown-induced tumor suppression. (B-C) Experimental setup to evaluate the effects of GPX4 knockdown combined with macrophage depletion on tumor growth, as assessed by tumor weight and representative images, in an orthotopic gastric cancer (GC) mouse model. (D-E) Design of the orthotopic mouse model used to assess the survival benefits of GPX4 knockdown and the impacts of macrophage depletion on survival outcomes. (F) Western blot analysis confirming the efficiency of GPX4 knockdown in the human gastric cancer cell line (NUGC3). (G) Schematic diagram of a design for investigating macrophage polarization changes and their functional impact. THP-1 cells, pretreated with 320 nM PMA for 6 h, were cultured with conditioned media from either vehicle or GPX4-knockdown GC cells for 48 h. (H) Flow cytometry analysis of macrophages following pretreatment with NUGC3 cell-conditioned medium. (I-L) Assays of NUGC3 cells co-cultured with macrophages including wound-healing assay (I), Transwell migration assay (J), colony formation assay (K) and apoptosis assay (L).

Figure 3

Metabolomic analysis of the impacts of GPX4 knockdown on metabolite levels and macrophage polarization. (A) Schematic representation of the metabolomic analysis workflow, including sample collection and the mass spectrometry procedure. (B) Quality control of metabolomic data, presented as principal component analysis (PCA) plots. (C) Volcano plot depicting differential metabolite families, including 5'-deoxyribonucleosides and azole carboxylic acids and derivatives, between vehicle and GPX4-knockdown NUGC3 cells. (D) Volcano plot highlighting differential metabolites, with a focus on N-methyl-kynurenine and kynurenine (kyn), in GPX4-knockdown NUGC3 cells. (E) Pathway enrichment analysis revealing significant enrichment in the tryptophan (Trp) metabolism pathway in GPX4-knockdown NUGC3 cells. (F) Flowchart of tryptophan and kynurenine metabolism. (G) Kynurenine levels, tryptophan levels, and the kynurenine/tryptophan ratio in the supernatant of GPX4-knockdown NUGC3 cells. (H-I) GPX4 expression and kynurenine levels in fresh gastric cancer tissues from 26 patients, including kynurenine and tryptophan measurements. (J) Flow cytometry analysis of macrophage polarization (M1 and M2) in response to varying kynurenine concentrations (50, 100, and 200 μM for 48 h) using THP-1 cells.

Figure 4

Regulation of kynurenine metabolism by GPX4 through modulation of intracellular ROS levels. (A) Western blot analysis of kynurenine metabolism-related enzymes (KYNU, IDO1, and IDO2) in NUGC3 with GPX4 knockdown or overexpression. (B) KYNU and IDO1 expressions in NUGC3 cells after Erastin (1 and 2 μM) and Ferrostatin-1(0.5 and 1 μM) treatment. (C-D) Intracellular ROS levels measured by flow cytometry to evaluate the impacts of these treatments on ROS levels in NUGC3 cells with GPX4 knockdown or overexpression, or treated with hydrogen peroxide (H₂O₂) and 4-hydroxy-2-nonenal (4HNE) for 6 h, or N-acetylcysteine (NAC) for 24 h. (E) Dose-dependent effects of H₂O₂ and 4HNE on KYNU, IDO1, and IDO2 expression in NUGC3 cells. (F) Dose-dependent effects of NAC and TEMPOL on KYNU, IDO1, and IDO2 expression in NUGC3 cells. (G) Effects of ROS scavenging (via NAC) on KYNU, IDO1, and IDO2 expression in cells with GPX4 knockdown. (H) Kynurenine levels in NUGC3 cell culture supernatants after Erastin and Ferrostatin-1 treatment. (I-J) Kynurenine and tryptophan levels in cell culture supernatants and lysates, along with the kynurenine/tryptophan (Kyn/Trp) ratios, following ROS modulation by H₂O₂, 4HNE, or NAC.

Figure 5

Regulation of IDO1 expression and kynurenine metabolism by GPX4 knockdown in gastric cancer cells. (A) Quantitative PCR analysis of the effect of GPX4 knockdown on IDO1 mRNA levels in NUGC3 cells. (B-C) Western blot analysis of the impacts of GPX4 knockdown on the expression of AhR, IL-6, and p-STAT3. (D-E) Effects of exogenous ROS on the expression of AhR, IL-6, and p-STAT3, as assessed by Western blotting. (F) Impacts of ROS scavenging (NAC treatment) on the expression of AhR, IL-6, and p-STAT3. (G) Western blot analysis of GPX4, IDO1, and KYNU protein expression levels to verify overexpression efficiency. (H) Quantification of kynurenine, tryptophan, and kynurenine/tryptophan (Kyn/Trp) ratio in the cell lines described in the panel. (I) Western blot analysis of protein expression levels of GPX4, KYNU, p-STAT3, and IDO1 to verify knockdown efficiency and regulation. (J) Quantification of kynurenine, tryptophan, and the kynurenine/tryptophan ratio in the cell lines described in the panel.

Figure 6

GPX4 knockdown regulates KYNU protein stability by altering ubiquitination patterns. (A) Quantitative PCR analysis of KYNU mRNA levels in NUGC3 cells following GPX4 knockdown. (B) Western blot analysis of KYNU protein expression in vehicle and GPX4-knockdown cells treated with the proteasome inhibitor MG132 (5 μM, 24 h) or the autophagy inhibitor BafA1 (10 μM, 24 h). (C-D) Degradation pathways of the KYNU protein in NUGC3 cells pre-treated with MG132 (5 μM) or BafA1 (10 μM) for 18 h, further co-treated with H₂O₂ (0.8 mM) or NAC (10 mM) for 6 h. (E) KYNU protein degradation over time, analyzed by western blot, in control and GPX4-knockdown NUGC3 cells treated with cycloheximide (CHX, 100 ng/mL) for 0, 4, 8, and 12 h. (F) Western blot analysis of pathways involved in KYNU degradation in NUGC3 cells treated with CHX (100 ng/mL) in combination with MG132 (5 μM) or BafA1 (10 μM) for 0, 6, and 12 h. Immunoprecipitation assays were conducted in (G-L). (G) In NUGC3 cells with GPX4 knockdown or transfected with the control vector, Flag-KYNU was overexpressed and treated with MG132 (5 μM) for 24 h. (H-I) In NUGC3 cells overexpressing Flag-KYNU, cells were treated with MG132 (5 μM) for 24 h. (J) NUGC3 cells overexpressing Flag-KYNU were also transfected with HA-ubi-K48 or HA-ubi-K63 constructs and treated with H₂O₂ (0.8 mM) and/or MG132 (5 μM). (K-L) In NUGC3 cells overexpressing Flag-KYNU and HA-ubi-K48 or HA-ubi-K63, cells were treated with H₂O₂ (0.8 mM) or NAC (10 mM) in combination with MG132 (5 μM).

Figure 7

Identification and functional analysis of the key ubiquitination sites of the KYNU protein modulated by ROS. (A) Potential ubiquitination sites of the KYNU protein were identified using http://www.phosphosite.org. (B) Six frequently modified lysine (K) residues of KYNU were selected from the literature and mutated to arginine (R). (C) Western blot analysis of KYNU stability in NUGC3 cells overexpressing wild-type or lysine-mutant KYNU (K96R, K163R, and other mutants) following treatment with NAC (10 mM) for 24 h. Results indicate that K96R mutation prevents NAC-induced degradation of KYNU. (D) Co-immunoprecipitation analysis of K48-ubiquitinated KYNU in NUGC3 cells overexpressing wild-type KYNU or K96R-KYNU after a 24-hour treatment with MG132 (5 μM) and NAC (10 mM). Results show a significant reduction in K48-ubiquitination in the K96R mutant. (E) Western blot analysis of KYNU stability in NUGC3 cells overexpressing wild-type or lysine-mutant KYNU (K96R, K163R, and other mutants) after treatment with H₂O₂ (10 mM) for 6 h. The K163R mutation prevents ROS-induced stabilization of KYNU. (F) Co-immunoprecipitation analysis of K63-ubiquitinated KYNU in NUGC3 cells overexpressing wild-type KYNU or K163R-KYNU following pre-treatment with MG132 (5 μM) for 18 h and further co-treatment with H₂O₂ (10 mM) for 6 h. The results indicate that the K163R mutation reduces K63-ubiquitination of KYNU. (G-H) Expression of flag-KYNU in WT-, K163R-, and K96R-KYNU plasmid-transfected control and GPX4 knockdown NUGC3 gastric cancer cells at 36- and 60-hour post-transfection. (I) Kynurenine levels in cells with KYNU-K96R or KYNU-K163R mutations under NAC or H₂O₂ conditions to evaluate the functional effects on kynurenine metabolism. (J-K) Design of the co-culture strategy and colony formation assay in NUGC3 cells overexpressing wild-type KYNU, K96R-KYNU, or K163R-KYNU after treatment with H₂O₂ for 6 h or NAC for 24 h, showing differential effects on cell growth.

Figure 8

Mechanism by which GPX4 in tumor cells regulates kynurenine metabolism via ROS, influencing macrophage polarization in the tumor microenvironment. This graph illustrates the mechanism by which GPX4 knockdown in tumor cells affects kynurenine metabolism through the regulation of ROS levels, ultimately affecting macrophage polarization within the tumor microenvironment. Briefly, GPX4 knockdown in tumor cells increases intracellular ROS levels, which alters the ubiquitination pattern of KYNU, promoting its stabilization. This results in reduced intracellular and extracellular kynurenine levels. The decreased availability of kynurenine in the tumor microenvironment reduces M2 macrophage polarization and enhances M1 macrophage polarization, leading to an overall antitumor effect.