Polyamine-Blocking Drug DFMO Weakens Mutant K-RAS Signaling in Pancreatic Cancer Cells

Pancreatic cancer remains one of the most difficult malignancies to treat, and a single genetic lesion sits at the heart of most cases: mutations in the K-RAS oncogene, which are found in more than 90% of human pancreatic adenocarcinomas. A new open-access study published in the journal Amino Acids reports that an older, repurposed drug called α-difluoromethylornithine, or DFMO, can push against that dependency from an unexpected direction. Rather than attacking the mutant K-RAS protein itself, DFMO starves pancreatic cancer cells of polyamines—small positively charged molecules that the tumors need in abundance—and in doing so, the study found, it dampens several of the signaling pathways that mutant K-RAS drives.

The research, conducted by Hwudaurw Chen, David E. Stringer, Haiyan Cui, Eugene W. Gerner and Natalia A. Ignatenko, set out to determine whether blocking polyamine synthesis could serve as a valid way to target the metabolic dependency of mutant K-RAS pancreatic tumors, in comparison with direct suppression of the oncogenic K-RAS protein. The team worked with the MIA PaCa-2 human pancreatic cancer cell line, which carries a homozygous K-RASG12C mutation, and with M-27, an isogenic clonally selected derivative of that line in which the K-RASG12C mutation was disrupted by homologous recombination. Because of that disruption, M-27 cells do not express a functional GTP-bound RAS protein, making them a laboratory stand-in for what happens when oncogenic RAS activity is eliminated—the molecular state that a new class of RAS (ON) inhibitors is designed to produce.

The choice of model matters. K-RAS mutations are the most common genetic feature of pancreatic ductal adenocarcinoma, and the study notes that the G12D variant accounts for roughly 40% of all cases and G12V for about 30%, while the G12C variant found in MIA PaCa-2 cells occurs in only 1–2% of patients. The amino acid substitutions in the mutant RAS protein impair GTP hydrolysis, allowing the active, GTP-bound form of the protein to accumulate in cancer cells and broadcast proliferation-stimulating signals through downstream effectors including the ERK1/2 kinases, RAC GTPases, AKT, PI3K lipid kinases and the Ral guanine exchange factors. Targeting that network directly has proven difficult: small-molecule inhibitors of mutant K-RASG12C, such as sotorasib and adagrasib, have shown mixed success, benefiting only patients with rare molecular subtypes and often failing because secondary oncogenic mutations in K-RAS and other genes confer drug resistance.

DFMO takes a different route into the same network. The drug irreversibly inhibits ornithine decarboxylase, the rate-limiting enzyme of polyamine biosynthesis, thereby depleting the cell’s supply of putrescine, spermidine and spermine. These polycations are essential for normal cellular growth as well as tumorigenesis, and the pancreas normally contains high amounts of them, especially spermidine. DFMO is already approved for the treatment of trypanosomiasis, hirsutism and neuroblastoma, and clinical studies in people at risk for colon, prostate and skin cancers have suggested it may help prevent those diseases. Elevated ornithine decarboxylase activity and high polyamine content have been reported in pancreatic cancer cell lines and in tumor samples from patients, and earlier work in a genetically engineered mouse model of pancreatic cancer showed that DFMO suppressed PanIN lesion progression driven by mutant K-RAS.

In the new experiments, the two cell lines turned out to be more similar than their genetic difference might suggest. Control MIA PaCa-2 and M-27 cells had comparable intracellular polyamine levels and grew at statistically indistinguishable rates, indicating that the polyamine metabolic pathway resists inhibition of mutant K-RAS GTPase activity. DFMO treatment, however, effectively suppressed growth and polyamine content in both lines. Putrescine and spermidine levels in treated MIA PaCa-2 cells fell by more than 10-fold on average starting on the first day after subculture, and total polyamine content was significantly reduced on days 3 and 4 (p < 0.01). In M-27 cells, the same metabolites dropped by more than 6-fold on day 1, with total polyamine suppression evident from day 2 onward. Cell viability, notably, was not significantly affected, meaning the drug slowed growth rather than simply killing the cells outright.

The signaling analysis revealed a selective pattern of change. In MIA PaCa-2 cells treated with 2.5 millimolar DFMO, the most pronounced effect was on the ERK1/2 pathway: the level of phosphorylated ERK1/2 protein fell by approximately 2-fold after 96 hours of treatment, reaching a degree of suppression comparable to that seen in the untreated M-27 cells, which lack active RAS. DFMO also reduced the protein level of the RAS-like small GTPase RalA—by 2-fold at 48 hours and more than 3-fold at 96 hours—while leaving its effector RalBP1 and the exocyst proteins EXO84 and SEC5 unchanged. In contrast, DFMO treatment slightly increased phosphorylated PI3K in MIA PaCa-2 cells and did not significantly alter phosphorylated AKT, a pattern the authors interpret as a possible compensatory response to the loss of oncogenic stimuli. Interestingly, DFMO induced c-MYC protein levels 3-fold in both cell lines at 96 hours, yet ornithine decarboxylase expression was not altered, showing that the drug inhibits the enzyme’s activity without changing its abundance.

Two further molecular findings connect polyamine depletion to the malignant behavior of the cells. First, DFMO treatment lowered the amount of K-RAS protein itself in the membrane fraction of MIA PaCa-2 cells in a time-dependent manner, with the most notable decrease—2-fold relative to the 96-hour control and 3.4-fold relative to the 48-hour control—observed with 2.5 millimolar DFMO at 96 hours; the authors cautiously suggest this could be mediated by let-7 family microRNAs, which are known to negatively regulate K-RAS expression and are upregulated when polyamines are depleted. Second, DFMO suppressed the level of hypusinated eIF5A by 2-fold. Hypusination is a modification that requires spermidine and is essential for the function of this translation initiation factor, which regulates proteins involved in cytoskeleton formation, migration and invasion in pancreatic cancer cells. Consistent with these changes, F-actin was visible in the protrusions of control cells but absent in cells treated with either 2.5 or 5 millimolar DFMO, the EMT-promoting transcription factor Snail was reduced 2-fold in treated MIA PaCa-2 cells, and DFMO as a single agent suppressed cell migration and invasion in transwell assays more effectively than the PI3K inhibitor LY294002.

In a SCID mouse orthotopic model, MIA PaCa-2 or M-27 cells were implanted into the pancreas, and one week later the animals received DFMO at 500 milligrams per kilogram per day by oral gavage—a dose the authors note is equivalent to one tested in human clinical trials. DFMO-treated MIA PaCa-2 tumor-bearing animals survived significantly longer than vehicle-treated controls, with a median survival of 53 days versus 36 days (p = 0.0184), a 52.80% increase in median survival time. Strikingly, that median survival matched that of mice implanted with the M-27 line, which lacks active RAS (MIA PaCa-2 control versus M-27 control, p = 0.0204). By contrast, DFMO did not significantly improve survival in mice bearing M-27 tumors (p = 0.6055), and the survival gain in that group was only 5.70%. The authors interpret this as evidence that DFMO suppresses pancreatic tumorigenesis driven by mutant K-RAS similarly to targeted elimination of the active GTP-bound RAS protein—and that once oncogenic RAS activity is already gone, polyamine blockade adds little benefit, possibly because M-27 cells have adapted through metabolic routes less sensitive to the drug. No significant weight changes were observed between treated and control animals during the experiment.

The study’s limitations are clearly drawn. The immunodeficient orthotopic model could not capture DFMO’s effects on anti-tumor immunity, although previous work in immunocompetent mouse models found that DFMO can reverse the immunosuppressive pancreatic tumor microenvironment by increasing infiltration of CD4-positive and CD8-positive T lymphocytes. The authors also note that metabolic pathways not evaluated here could be modulated by the drug, and that the heterogeneity and adaptability of pancreatic tumors may challenge even the newest RAS (ON) inhibitors, such as daraxonrasib, which recently showed significantly longer overall and progression-free survival than chemotherapy in a Phase 3 trial of previously treated metastatic PDAC patients. With the polyamine analog SBP-101 already under clinical investigation in combination with nab-paclitaxel and gemcitabine in newly diagnosed PDAC patients, the findings add to the rationale for targeting polyamine metabolism—alone or in combination with other drugs—in tumors that remain addicted to activated K-RAS, including those that have developed resistance to direct mutant K-RAS inhibitors.

Subject of Research: Effects of the polyamine synthesis inhibitor DFMO on mutant K-RAS signaling and tumor growth in pancreatic cancer cell and mouse models

Article Title: Difluoromethylornithine modulates mutant K-RAS signaling pathways in pancreatic cancer cells

Article References: Chen, H., Stringer, D. E., Cui, H., Gerner, E. W., & Ignatenko, N. A. (2026). Difluoromethylornithine modulates mutant K-RAS signaling pathways in pancreatic cancer cells. Amino Acids, 58(1), Article 60. https://doi.org/10.1007/s00726-026-03559-5

Image Credits: AI Generated

DOI: 10.1007/s00726-026-03559-5

Keywords: pancreatic cancer, DFMO, K-RAS, polyamine metabolism, ornithine decarboxylase, ERK1/2, RalA, eIF5A hypusination, orthotopic mouse model, MIA PaCa-2, targeted therapy, PDAC