GPX1 and OSBPL8 team up at the ER to shield cancer cells from ferroptosis
Researchers find GPX1 is recruited to the ER by OSBPL8 to block ROS-driven ferroptosis; disrupting either one slows tumor growth in mouse models.
Summary
Ferroptosis, an iron-dependent form of cell death driven by oxidized lipids, can suppress tumors even without drugs like GPX4 inhibitors. This study shows how cancer cells resist this natural, ROS-triggered form of ferroptosis. A genome-wide CRISPR screen identified GPX1 as a specific brake on ROS-induced ferroptosis, but not on ferroptosis caused by the GPX4 inhibitor RSL-3. GPX1 works with OSBPL8, an ER-anchored lipid-binding protein that recruits it to the endoplasmic reticulum, where lipid peroxidation builds up before the cell membrane ruptures. Knocking out GPX1 or OSBPL8 sensitized melanoma, lung and pancreatic cancer cells to ROS-induced ferroptosis and slowed xenograft tumor growth, an effect reversed by the ferroptosis inhibitor ferrostatin-1. The GPX1–OSBPL8 axis is a possible new cancer target.
Detailed Summary
Ferroptosis is a form of iron-dependent, non-apoptotic cell death caused by lethal build-up of oxidized membrane lipids. Most work has relied on artificial triggers such as GPX4 inhibitors (RSL-3) or GPX4 deletion. How ferroptosis happens naturally in living tumors, where nothing is inducing it, is much less clear. Cancer cells run at high ROS levels because of their altered metabolism, which may make them vulnerable to a noncanonical, ROS-induced ferroptosis. Earlier work from this group tied that process to peroxidation of phosphatidic acid (PA) rather than phosphatidylethanolamine, and showed it does not depend on GPX4 or ACSL4.
The team ran a genome-wide CRISPR-Cas9 knockout screen (18,436 genes) in A375 melanoma cells treated with the ROS generator tert-butyl hydroperoxide (TBH). The screen recovered known regulators (SLC7A11, PHLDA2, GPAT3, LRP8, NFS1) and also flagged GPX1. Knocking out GPX1 in A375, A549 lung, and MIAPaCa2 and S2-013 pancreatic cancer cells increased TBH-induced death and lipid peroxidation but had no obvious effect on RSL-3-induced death. The death was rescued by ferrostatin-1 and deferoxamine, but not by necroptosis or apoptosis inhibitors. Re-expressing GPX1 reversed the sensitivity. GPX1 loss did not alter H2O2-induced necrosis.
Affinity purification and mass spectrometry identified OSBPL8, an ER-anchored lipid-binding protein, as a major GPX1 partner. The two interact through OSBPL8's N-terminal domain (residues 1–147) and a region of GPX1 around residues 130–137, which is distant from the Sec49 catalytic site. OSBPL8 knockdown or knockout, like GPX1 loss, sensitized cells to ROS-induced but not RSL-3-induced ferroptosis. In double-knockout cells, only co-expression of both proteins gave full rescue. A GPX1 Δ130–137 mutant still lowered ROS but protected poorly against lipid peroxidation and cell death, so the benefit goes beyond general antioxidant activity.
Live-cell BODIPY C11 imaging showed lipid peroxidation accumulating and concentrating at the ER, near the plasma membrane, over about 7 hours before blebbing and rupture. OSBPL8 recruited GPX1 to the ER, and OSBPL8 lacking the N-terminal or transmembrane domain failed to suppress ER lipid peroxidation. The text provided is truncated before the PA-peroxidation mechanism and the later in vivo OSBPL8 data, though the abstract states that GPX1 directly reduces oxidized PA.
In mice, GPX1-null A375, A549 and S2-013 xenografts grew less than controls without any treatment. They showed higher ferroptosis markers (hyperoxidized PRDX3, PTGS2, CHAC1) and more lipid peroxidation, and ferrostatin-1 largely reversed the growth suppression. GPX1 is overexpressed in several cancers, and high expression correlates with poorer survival in pancreatic adenocarcinoma. Together, the results point to the GPX1–OSBPL8 axis as a potential therapeutic vulnerability. Caveats include reliance on xenograft models and cell lines, and the use of TBH as the ROS source.
Key Findings
- A CRISPR screen identified GPX1 as a specific suppressor of ROS-induced ferroptosis, with no effect on RSL-3 (GPX4 inhibitor) induced death.
- OSBPL8 binds GPX1 via its N-terminal domain and recruits GPX1 to the ER, where ROS-driven lipid peroxidation accumulates.
- A GPX1 Δ130–137 mutant lost OSBPL8 binding and protected poorly against lipid peroxidation despite still lowering ROS.
- GPX1 or OSBPL8 loss sensitized melanoma, lung and pancreatic cancer cells to ferroptosis across multiple cell lines.
- GPX1-null xenografts grew less without any drug, and ferrostatin-1 largely reversed this, supporting ferroptosis as the mechanism.
Methodology
Genome-wide CRISPR-Cas9 knockout screen in A375 cells under TBH stress, followed by knockout, knockdown and rescue experiments in several cancer cell lines. Protein interactions were mapped by affinity purification/mass spectrometry, co-IP and deletion mutants, and localization by C11-BODIPY and immunofluorescence imaging. Tumor effects were tested in xenograft mouse models, with and without ferrostatin-1.
Study Limitations
Evidence comes from cultured cancer cell lines and immunodeficient xenograft models, mostly using TBH as the ROS source, so generalizability to patients is uncertain. The provided full text is truncated, so the PA-peroxidation mechanism and the in vivo OSBPL8 data could not be fully assessed. Systemic effects of inhibiting GPX1 and OSBPL8 in normal tissues were not addressed in the available text.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: