Ferroptosis is a relatively recent addition to the family of regulated cell death pathways, and it differs fundamentally from apoptosis, the form of cell death most traditional therapies aim to trigger. Instead of orderly cellular dismantling, ferroptosis is a violent, iron-driven process in which lipid membranes are destroyed by peroxidation—essentially, the cell’s fatty boundaries are oxidized until they rupture. Because cancer cells often possess elevated iron stores and heightened baseline oxidative stress, researchers have long hypothesized that they might be exquisitely sensitive to this form of death. Yet pancreatic tumors have proven resistant even to ferroptosis-inducing compounds, and the Verona team set out to discover why.

Their central suspect was p53, the famous “guardian of the genome.” In healthy cells, p53 acts as a tumor suppressor, halting cell division and initiating death programs when damage is detected. But in pancreatic cancer, the TP53 gene is frequently mutated—and, remarkably, many of these mutations do not merely disable the protein. Instead, they endow it with new, gain-of-function activities that actively promote tumor survival, metastasis, and therapy resistance. Whether mutant p53 influenced ferroptosis susceptibility had remained poorly understood, and the answer, it turns out, is emphatically yes.

To isolate the effect, the researchers employed isogenic pancreatic cancer cell models—cell lines that are genetically identical except for the status of TP53. Using CRISPR-Cas9 gene-editing technology, they created TP53 knockout cells in which the mutant gene was deleted entirely, and they also transiently overexpressed common mutant TP53 variants in these backgrounds. The comparison proved striking. When mutant TP53 was removed, pancreatic cancer cells became dramatically more vulnerable to ferroptosis. They accumulated more reactive oxygen species, suffered greater lipid peroxidation—as measured by markers including 4-hydroxynonenal—and displayed clear signs of mitochondrial dysfunction. The deletion of mutant p53, in other words, stripped away a protective shield, leaving the cells exposed to the oxidative assault that defines ferroptotic death.

Conversely, cells expressing mutant p53 preserved the structural integrity of their mitochondria under ferroptotic stress, sustaining their bioenergetic flexibility even as the lethal insult pressed in. Transmission electron microscopy and measurements of mitochondrial membrane potential confirmed that these cells kept their power factories functional where their p53-deficient counterparts faltered. This mitochondrial preservation was not a passive trait but the visible outcome of an elaborate adaptive program that mutant p53 orchestrates at the transcriptional level.

To map that program, the researchers turned to RNA sequencing, profiling the full complement of gene expression changes triggered by ferroptosis inducers in cells with and without mutant p53. The transcriptomic analysis revealed a multi-layered defensive network. Mutant p53-expressing cells ramped up antioxidant genes that neutralize the reactive oxygen species driving lipid peroxidation, and simultaneously activated a suite of metabolic genes. Among the most significant pathways to emerge was PI3K–AKT signaling, a pro-survival cascade that in these cells was linked to a selective shift toward glycolysis—the fermentation of glucose into lactate—as a means of maintaining cellular ATP, the universal energy currency.

That metabolic pivot proved to be the linchpin of the resistance. Using metabolic flux assays, the team measured both glycolytic activity and mitochondrial respiration, quantified through the oxygen consumption rate, and found that mutant p53-expressing cells could flexibly toggle between oxidative phosphorylation and glycolysis to keep their energy supply steady under stress. The functional consequences were demonstrated directly: supplementing the culture medium with extra glucose enhanced the survival of mutant TP53 cells treated with ferroptosis inducers, while blocking glycolysis with inhibitors such as 2-deoxy-D-glucose impaired their survival. Critically, neither manipulation had the same effect in TP53-knockout cells, confirming that the glycolytic lifeline exists only where mutant p53 is present.

The most clinically significant portion of the study came next. Rather than attacking the metabolic adaptation alone, the researchers tested whether pharmacological reactivation of wild-type p53 could collapse the entire defensive network. They used APR-246, also known as eprenetapopt, a small molecule designed to restore wild-type conformation and function to mutant p53, in combination with ferroptosis inducers including imidazole ketone erastin, a well-characterized inhibitor of the cystine transporter that fuels the antioxidant machinery of cells. The combination proved devastating to the tumor cells. Reactivating wild-type p53 disrupted the adaptive transcriptional and metabolic program, abrogated the glycolytic reprogramming that had sustained ATP production, and significantly increased ferroptotic cell death.

Importantly, this effect was not confined to laboratory dishes. The researchers validated their findings in orthotopic murine models—in which pancreatic tumors are established in the pancreas of living animals, recreating the tumor microenvironment far more faithfully than cell culture. In these models, the combination of p53 reactivation and ferroptosis induction significantly increased tumor cell death, providing in vivo evidence that the strategy could translate beyond the petri dish. The work also benefited from mouse KPC-derived cell lines, 7940Bb and MT3, derived from genetically engineered mouse models of pancreatic cancer and provided through collaborations with Cold Spring Harbor Laboratory and the University of Pennsylvania.

The implications for treatment are considerable. Pancreatic ductal adenocarcinoma is characterized by late diagnosis, rapid progression, and profound resistance to chemotherapy, radiotherapy, and the targeted agents that have transformed outcomes in other cancers. Most patients survive only months after diagnosis, and the five-year survival rate remains among the lowest of any major cancer. Ferroptosis induction has been proposed as a way around this resistance precisely because it targets vulnerabilities—iron metabolism, lipid repair, antioxidant defense—that conventional therapies ignore. But the Verona study demonstrates that pancreatic tumors are not passive targets: mutant p53 endows them with a metabolically flexible, transcriptionally orchestrated armor that must be breached for ferroptosis to succeed.

The study also adds a new dimension to the biology of mutant p53 gain-of-function. Rather than simply evading apoptosis or promoting proliferation, mutant p53 here acts as a metabolic arbiter, rewiring how cells produce and protect energy so that lethal lipid peroxidation can be withstood. It links three of the hottest themes in modern cancer research—p53 biology, ferroptosis, and metabolic plasticity—into a single mechanistic framework. The finding that PI3K–AKT signaling couples p53 mutation to glycolytic shift suggests additional pharmacological points of intervention; the researchers demonstrated that glycolytic inhibition and PI3K pathway blockade each undermined the survival advantage of mutant TP53 cells under ferroptotic stress.

The authors are careful to frame the work as preclinical, and substantial hurdles remain before a combination of eprenetapopt and ferroptosis inducers reaches the clinic. Eprenetapopt itself has had a mixed record in clinical trials for blood cancers, and questions of drug delivery to the dense, poorly vascularized pancreatic tumor microenvironment remain formidable. Still, the identification of a therapeutically actionable vulnerability—one that can be pharmacologically flipped—offers a rare piece of good news in a disease that has seen painfully few. The work was supported by the Italian Association for Cancer Research, the Italian Ministry of University and Research, and European recovery fund programs, reflecting sustained investment in pancreatic cancer metabolism research at Verona.

What makes the study resonate beyond pancreatic cancer is the broader principle it establishes: cell death pathways do not operate in isolation from tumor metabolism, and the mutated genes that drive cancer also decide which death programs remain available. For the many tumors that harbor TP53 mutations, the ability of restored wild-type p53 to sensitize cells to ferroptosis suggests a generalizable combination strategy. For patients with pancreatic ductal adenocarcinoma—a disease desperate for options—the demonstration that a drug pair can strip away a tumor’s metabolic armor and ignite ferroptosis from within represents exactly the kind of mechanistic insight from which the next generation of therapies may be built.

The investigation centered on comparing two genetically defined cancer cell cultures derived from models of NSCLC: the ATF4-high KP-O and the ATF4-low KP-Y populations. Initial drug screening revealed a surprising specificity in their metabolic dependencies. While both were unresponsive to inhibitors targeting various amino acid transporters and metabolic enzymes such as SLC7A11, BCAT, or PHGDH, KP-O cultures exhibited heightened sensitivity to glutamine deprivation and treatment with DON, a glutamine analogue toxic to cancer cells. This disparity underscores the pivotal role of glutamine metabolism in determining therapeutic responses.

Delving deeper, the focus shifted towards targeting glutaminolysis—the conversion of glutamine to glutamate—central to cellular bioenergetics and biosynthesis. The study employed glutaminase inhibitors (GLSi) CB-839 (telaglenastat) and BPTES, agents capable of halting this rate-limiting step. Remarkably, KP-O cells demonstrated pronounced sensitivity to both drugs, linking high ATF4 expression and glutaminolysis dependency. Additionally, antagonism of ASCT2, the principal glutamine transporter using V-9302, mirrored this effect, further accentuating glutamine’s indispensable role in sustaining KP-O cell viability.

Mechanistic exploration through metabolic rescue experiments painted a nuanced picture of glutamate’s centrality. Pretreatment of KP-O cultures with dimethyl-2-oxoglutarate (DMG), a cell-permeable α-ketoglutarate analog that replenishes critical TCA cycle intermediates, or pyruvate derived from glucose metabolism, effectively reversed sensitivity to CB-839. This rescue was unique as other tested metabolites or antioxidants failed to confer protection, except for erastin, a cysteine–glutamate antiporter system inhibitor. These findings emphatically pinpointed glutamate exhaustion rather than downstream metabolic disruptions as the culprit for GLSi-induced cytotoxicity in KP-O cultures.

Critical to the narrative is the integral role played by ATF4. Genetic ablation or pharmacological attenuation of ATF4 activity using ISRIB (Integrated Stress Response Inhibitor) rendered KP-O cells resistant to CB-839, underscoring the dependency of glutaminase sensitivity on this transcription factor. Conversely, forced ATF4 overexpression in the historically resistant KP-Y cultures conferred newfound vulnerability to GLS inhibition, demonstrating a causal relationship. This interplay also extended to 3D tumor spheroid models: KP-O spheroids lost their characteristic anoikis resistance—a hallmark of metastatic potential—upon GLSi or V-9302 treatment, a defect that was likewise reversed when ATF4 was inhibited.

From a translational perspective, in vivo experiments confirmed the therapeutic promise of targeting glutaminolysis within the metastatic microenvironment. Intravenous transplantation of KP-O cultures into murine hosts resulted in aggressive lung metastasis under vehicle treatment but was nearly abolished with CB-839 administration. Strikingly, KP-Y cells implanted similarly evoked minimal metastatic burden regardless of treatment, demonstrating specificity. Notably, CB-839 did not impede the primary tumor growth in either model following subcutaneous transplantation, a divergence highlighting the metastasis-focused effectiveness of GLS inhibition.

Quantitative assessments endorsed these observations, with CB-839 treatment virtually eradicating distant metastases from KP-O tumors without affecting their primary mass or growth kinetics. This selective suppression of metastatic seeding or outgrowth, sparing tumor proliferation, suggests a unique dependency of metastatic cells on glutaminolysis mediated by ageing and ATF4 activation. These insights may explain the clinical challenges in treating metastasis and underscore the need for tailored metabolic interventions targeting this axis.

This study pioneers the conceptual junction where ageing biology intersects with cancer metabolism and metastasis. The integrated stress response, governed by ATF4, commandeers metabolic rewiring that fosters metastatic competence through glutamine and glutamate utilization. By exploiting this axis using clinically relevant GLS inhibitors, such as CB-839, there appears to be a viable strategy to thwart metastasis specifically in cancers with elevated ATF4 signaling—a phenotype enriched in aged patients.

Future clinical translation of these findings could revolutionize NSCLC management in older demographics, where current therapies exhibit limited efficacy against metastatic disease. It reveals how stress-adaptive transcription factors reshape metabolic landscapes within tumors, creating transient but exploitable vulnerabilities. Moreover, it invites broader applications across cancers exhibiting stress response hyperactivation, potentially heralding a new class of metabolically targeted anti-metastatic agents.

In summary, the novel identification of ageing-induced ATF4-dependent glutamine addiction in metastatic NSCLC cells presents a compelling target for intervention. GLS inhibitors, currently progressing through clinical trials, may find renewed focus as adjuvants to prevent metastatic progression rather than solely tumor reduction. This paradigm shift champions metabolic stress signaling as the Achilles’ heel of metastatic dissemination, reshaping therapeutic paradigms in oncology.

The study’s rigorous integration of cellular, molecular, and in vivo models highlights the precision with which cancer metabolism can be therapeutically manipulated. The metabolic plasticity modulated by ATF4 not only sustains metastasis but unveils a highly selective, context-dependent vulnerability. In doing so, it sets a precedent for unraveling complex age-related oncogenic programs through metabolic intervention, promising enhanced survivorship and quality of life for patients burdened by aggressive lung cancers.

As this research lays the groundwork for targeted metabolic therapies, it also prompts vital questions about long-term effects, resistance mechanisms, and patient stratification. The intersection of ageing biology with cancer therapeutics will undoubtedly continue to burgeon, catalyzing innovative strategies that are as complex and adaptive as the disease they aim to conquer.

Subject of Research: Metabolic rewiring driven by ATF4 in ageing and its impact on metastasis in non-small cell lung cancer.

Article Title: Ageing promotes metastasis via activation of the integrated stress response.

Article References:
Patel, A.A.H., Dzanan, J.J., Ali, K.X. et al. Ageing promotes metastasis via activation of the integrated stress response. Nature (2026). https://doi.org/10.1038/s41586-026-10216-0

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41586-026-10216-0