Dying cancer cells build actin scaffolds inside their nuclei as they turn acidic during ferroptosis
In ferroptosis-sensitive cancer cells, nuclear actin filaments form as cells acidify, and alkaline conditions block them. They appear to be a side effect of ferroptosis, not a cause.
Summary
Ferroptosis is an iron-dependent form of cell death that is being studied in cancer, Parkinson's disease, and ischemia-reperfusion injury. Researchers exposed human fibrosarcoma cells (HT-1080) to the ferroptosis-inducing drug RSL3 and found that filaments of actin, a structural protein, assemble inside the nucleus. This appeared about 90 minutes after treatment. The filaments formed as actin was imported into the nucleus through a transporter called IPO9, while the actin scaffold in the cytoplasm broke down. Blocking ferroptosis with ferrostatin-1 or deferoxamine prevented the filaments. Making the surrounding fluid more alkaline also prevented them, and the cells' interior grew steadily more acidic as ferroptosis proceeded. Disrupting nuclear actin did not change how quickly cells died, so the filaments look like a side effect of ferroptosis, not a driver. This is a cell-culture study in a single cancer cell line.
Detailed Summary
Ferroptosis is an iron-dependent form of programmed cell death driven by lipid peroxidation and membrane rupture. It is implicated in tumor biology, Parkinson's disease, ischemia-reperfusion injury, and skin disorders. Researchers are trying to induce it in resistant tumors and block it in degenerating tissue, yet its mechanics are incompletely mapped. Actin filaments inside the nucleus, once thought not to exist, are now linked to DNA repair, chromatin organization, and transcription. This study asked whether nuclear F-actin appears during ferroptosis.
The team used ferroptosis-sensitive HT-1080 fibrosarcoma cells treated with the GPX4 inhibitor RSL3 (0.5-2 μM). Nuclear F-actin was detected with phalloidin, SiR-actin, and anti-actin antibodies. The authors also built a stable line expressing a nuclear actin chromobody (nAC-TagGFP2) for live 3D and time-lapse imaging. Cell death was scored by the ratio of SYTOX Green-positive to Hoechst-positive nuclei. All experiments used three biological replicates, with t-tests or ANOVA and data shown as mean ± SEM.
Untreated cells showed diffuse nuclear signal with no prominent bundles. After RSL3, about 35% of cells in serum-containing DMEM assembled nuclear F-actin bundles. In a 25-cell time analysis, bundles appeared roughly 90 minutes after treatment and persisted another ~90 minutes. They disassembled as the nuclear and plasma membranes ruptured. This is much longer than the transient nuclear F-actin described in earlier work. 3D reconstruction showed thin filaments near the nuclear periphery and thicker bundles in the nucleoplasm, mostly near the ventral side. Some clumps excluded Hoechst staining, hinting at chromatin redistribution. Ferrostatin-1 and deferoxamine markedly reduced the filaments, linking them to ferroptosis itself.
On mechanism, RSL3 raised nuclear actin and G-actin signals, as well as total cellular actin signal. Knocking down the actin importer IPO9 markedly cut nuclear F-actin, pointing to an import-dependent process. Cytoplasmic F-actin first rose briefly, then progressively disassembled. Latrunculin A, which depolymerizes cytoplasmic filaments, increased nuclear F-actin, suggesting an inverse coupling between the two pools.
The nuclear filaments appear not to affect cell fate. Nuclear-targeted wild-type, polymerization-defective (R62D), and filament-stabilizing (S14C) actin changed nuclear assembly but caused no significant differences in RSL3-induced death. A cytoplasmic R62D mutant slightly delayed ferroptosis, while cytoplasmic S14C had no significant effect. This suggests cytoplasmic remodeling has a modest influence on timing, whereas nuclear F-actin is a concomitant feature.
The pH findings were unexpected. Nuclear F-actin was more frequent in PBS than in DMEM and least frequent in EBSS, the buffer with the most bicarbonate (26 mM, versus 4.2 mM in HBSS and 0 mM in PBS). Adding NaHCO3 to PBS abolished long bundles, and bicarbonate-free L-15 medium restored them. Raising extracellular pH with NaOH also suppressed filaments, pointing to pH, not bicarbonate itself. A mOrange2 reporter showed intracellular pH tracking extracellular pH and acidifying progressively during ferroptosis.
Caveats are substantial. Everything was done in one cancer cell line, using one inducer (RSL3), and the full text supplied here ends before the Discussion. The work is correlative on pH, and the function of the filaments is unknown. It offers no direct clinical or lifespan implications yet, but it adds a new cell-biological readout of ferroptosis.
Key Findings
- About 35% of HT-1080 cells assembled nuclear F-actin bundles during RSL3-induced ferroptosis in serum-containing DMEM; untreated cells showed no prominent bundles.
- Nuclear F-actin appeared ~90 min after RSL3 treatment and persisted ~90 more minutes (25-cell time analysis), disassembling at nuclear and plasma membrane rupture.
- Ferrostatin-1 (2 μM) and deferoxamine (100 μM) markedly inhibited nuclear F-actin, tying the structures to ferroptosis itself.
- IPO9 knockdown markedly reduced nuclear F-actin, and nuclear G-actin and total actin signals rose, indicating import-dependent assembly.
- Nuclear-targeted actin mutants (WT, R62D, S14C) showed no significant difference in RSL3-induced cell death, so nuclear F-actin appears concomitant, not causative.
- Cytoplasmic polymerization-defective actin R62D slightly delayed ferroptosis, while S14C had no significant effect.
- Nuclear F-actin fell as bicarbonate rose (0 mM PBS, 4.2 mM HBSS, 26 mM EBSS); NaOH also suppressed it, and a mOrange2 reporter showed progressive intracellular acidification.
Methodology
This was an in vitro cell-biology study in human HT-1080 fibrosarcoma cells, with ferroptosis induced by RSL3 and blocked by ferrostatin-1 or deferoxamine. Nuclear actin was visualized by phalloidin, SiR-actin, anti-actin staining, and a stable nuclear actin chromobody (nAC-TagGFP2) with live 3D and time-lapse confocal imaging. Mechanisms were probed with IPO9 siRNA, NLS-targeted and cytoplasmic β-actin mutants (R62D, S14C), latrunculin A, varied buffers, NaHCO3/NaOH, and a mOrange2 intracellular pH reporter read by flow cytometry (≥10,000 cells per well). Cell death was quantified by SYTOX Green/Hoechst ratio; stats used t-tests and ANOVA (mean ± SEM) with 3 biological replicates.
Study Limitations
Experiments used a single cancer cell line (HT-1080), mainly one ferroptosis inducer (RSL3), and 3 biological replicates, so generalizability to other cells, inducers, or in vivo tissue is untested. Nuclear F-actin did not alter ferroptotic sensitivity, so its function remains unknown, and the pH link is correlative (intracellular acidification favors, but is not proven to cause, assembly). The full text provided ended before the Discussion, so author-stated limitations and any conflict-of-interest disclosures could not be reviewed.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: