Cancer cells have a survival trick that nobody expected: they co-opt vitamin B2, one of the most ordinary nutrients in the human diet, to shield themselves from a form of cell death that would otherwise destroy them. Two independent research teams, working on separate continents, have now converged on the same finding. Riboflavin metabolism keeps a critical anti-death protein intact inside tumors, and an experimental compound called roseoflavin can strip that protection away, leaving cancer cells exposed.
The results, published in May 2026 in Nature Cell Biology and Nature Structural and Molecular Biology, reframe a common vitamin as a potential weak point in tumor biology and open a new front in the effort to kill cancer cells through a process called ferroptosis.
A lethal process cancer cells have learned to dodge
Ferroptosis is a form of iron-driven cell death that works by oxidizing the fatty lipids in cell membranes until the membrane collapses. Healthy cells defend against it through two parallel systems. The first relies on an enzyme called GPX4 and the antioxidant glutathione. The second runs through a protein called FSP1, which regenerates a lipid-soluble antioxidant called ubiquinol from coenzyme Q10.
Two landmark 2019 papers in Nature, one from the Bersuker group and another from the Doll and Olzmann-Dixon collaboration, independently established FSP1 as a glutathione-independent ferroptosis suppressor that works alongside GPX4. That discovery raised an urgent question: when oncologists block GPX4, many tumors simply lean on FSP1 to survive. What keeps FSP1 active?
Two labs, one answer: vitamin B2
The new studies answer that question with striking precision. A team led by Jose Pedro Friedmann Angeli at the University of Wurzburg, with PhD student Vera Skafar as a lead contributor, ran a focused CRISPR-Cas9 genetic screen and identified riboflavin, the active form of vitamin B2, as a direct modulator of FSP1 function. When cells had adequate riboflavin, FSP1 remained stable and active, blocking ferroptosis. When riboflavin was depleted or its metabolism disrupted, FSP1 protein levels dropped and cancer cells became far more sensitive to ferroptosis-inducing compounds.
“We are essentially targeting vitamin B2 metabolism to weaken the ferroptosis defense of cancer cells,” Friedmann Angeli’s group explained in a University of Wurzburg statement, describing the approach as turning an everyday nutrient into a druggable vulnerability rather than a blanket dietary concern.
Independently, the Olzmann group at UC Berkeley arrived at the same conclusion through a separate CRISPR screen using an endogenously tagged dual-fluorescent FSP1 reporter. That team pinpointed two specific enzymes, RFK and FLAD1, as the critical links between vitamin B2 and FSP1 stability. RFK converts riboflavin into flavin mononucleotide (FMN), and FLAD1 converts FMN into flavin adenine dinucleotide (FAD). Without either enzyme, FSP1 lost its structural integrity and its ability to suppress ferroptosis.
The convergence matters. Single-lab findings in cancer biology frequently fail to replicate, so having two independent groups confirm the same molecular mechanism at the point of initial publication substantially raises confidence in the core result.
Roseoflavin: the compound that exploits the weakness
The practical test came from roseoflavin, a naturally occurring riboflavin antimetabolite produced by the bacterium Streptomyces davawensis. When researchers treated cancer cells with roseoflavin, it disrupted the flavin supply chain, destabilized FSP1, and left cells exposed to ferroptosis triggered by a single compound.
Across multiple cancer cell lines, the pattern held: block the vitamin B2 pathway, and the FSP1 shield collapses. Cells that had previously resisted ferroptosis-inducing agents became vulnerable.
What the studies have not yet shown
Both studies relied primarily on cell-line experiments and, where animal data were included, mouse xenograft models. Neither team has published matched human tumor biopsies showing FSP1 protein or flavin levels before and after riboflavin modulation. That gap is significant because tumors in living patients exist in a metabolic environment far more complex than a culture dish. Blood riboflavin levels fluctuate with diet, gut absorption, and liver processing, and it remains unclear how much local flavin depletion a tumor would experience at a tolerable dose of roseoflavin.
Long-term safety data for roseoflavin in non-tumor tissues are also absent. Riboflavin is not a luxury nutrient. It serves as a cofactor for dozens of flavoproteins throughout the body, supporting mitochondrial energy production, antioxidant recycling, and amino acid metabolism. Blocking riboflavin metabolism broadly could, in theory, destabilize healthy tissues that depend on the same flavoprotein network. No large-animal toxicology study or pharmacokinetic profile for roseoflavin, alone or in combination with existing GPX4 inhibitors, has been published.
The therapeutic index, the margin between harming tumors and sparing normal cells, remains the central open question. One plausible hypothesis is that tumors with high expression of riboflavin transporters would be especially sensitive to roseoflavin while normal tissues compensate through alternative uptake routes. But that idea has not been tested in patient-derived organoids or clinical samples.
There is also a broader mechanistic caveat. FSP1 is not the only flavoprotein relevant to ferroptosis. Multiple enzymes involved in lipid metabolism, redox cycling, and mitochondrial respiration also rely on FAD or FMN. Disrupting riboflavin pathways, whether genetically or pharmacologically, could alter ferroptosis sensitivity through mechanisms that extend beyond FSP1. The current data support a primary role for FSP1, but they do not fully rule out additional flavoprotein contributions.
Where ferroptosis-based therapy stands more broadly
These findings arrive at a moment when ferroptosis is attracting serious attention from drug developers. Several groups have been pursuing GPX4 inhibitors and other ferroptosis-inducing strategies in preclinical pipelines, motivated by evidence that therapy-resistant cancer cells, particularly those that have undergone epithelial-to-mesenchymal transition, are often more vulnerable to ferroptosis than to conventional chemotherapy. The challenge has been that tumors frequently compensate by upregulating backup defenses like FSP1.
The riboflavin-FSP1 connection offers a potential way around that resistance. If roseoflavin or a more selective analogue could be paired with a GPX4-targeting agent, both arms of the ferroptosis defense would be compromised simultaneously. That combination logic is compelling on paper, but no published data yet show it working safely in a living organism.
What this does not mean for your diet
These studies do not justify restricting dietary vitamin B2 for people with cancer or at risk of cancer. The experiments manipulated riboflavin metabolism inside cells using genetic tools and pharmacologic doses of roseoflavin, not modest shifts in food intake. Clinical nutrition guidelines already recommend maintaining adequate intake of B vitamins to support overall health, and nothing in the current data overturns that advice.
The most realistic near-term implication is for drug development. Medicinal chemists may now try to design roseoflavin analogues or other small molecules that more selectively target the RFK-FLAD1-FSP1 axis in tumors while sparing essential flavoproteins in normal tissues. Any such program would require careful preclinical evaluation, including toxicology, metabolism, and off-target profiling, before reaching human trials.
For patients and oncologists, the key takeaway is conceptual rather than immediately actionable: common nutrients can become context-dependent vulnerabilities in cancer cells, especially when those cells rely on specific metabolic detours to survive therapy. Riboflavin metabolism now joins a growing list of such dependencies. Whether that insight will translate into a safe, effective ferroptosis-based treatment remains to be determined, but the mechanistic groundwork laid by these two converging studies is notably stronger than what most early-stage cancer targets can claim.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
