The concept of drug repurposing, sometimes called repositioning, has become an increasingly attractive strategy in oncology because established medicines already possess well-characterized safety profiles, known pharmacokinetics, and manufacturing pathways that would otherwise take years and substantial investment to develop from scratch. Antimalarial compounds in particular have drawn attention from cancer researchers because several of them, including chloroquine and its derivatives, have been shown to accumulate in acidic cellular compartments and to interfere with autophagy, a process that many tumors rely on for survival under metabolic stress. Mefloquine, a lipophilic quinoline developed in the 1970s for chloroquine-resistant malaria, crosses cell membranes readily and concentrates in lysosomes and mitochondria, properties that plausibly underlie the mitochondrial phenotype described in this study. Prior preclinical reports have suggested antitumor activity of mefloquine in models of glioma, leukemia, and other malignancies, but a systematic connection between mefloquine and ferroptosis in colorectal cancer had not been clearly established before this work.

Ferroptosis itself is a relatively recent addition to the catalog of regulated cell death modalities, first named in 2012. Unlike apoptosis, ferroptosis does not involve cell shrinkage, membrane blebbing, or caspase activation; instead it is defined by iron-dependent accumulation of lipid peroxides to levels that rupture the plasma membrane. The pathway is tightly governed by the balance between production of lipid reactive oxygen species through the Fenton reaction and enzymatic oxidation of membrane phospholipids, and the cellular defense system centered on glutathione peroxidase 4, which reduces lipid hydroperoxides using glutathione as a cofactor. Because rapidly dividing cancer cells already operate under elevated oxidative stress and often carry perturbations in iron metabolism, many investigators have proposed that pushing cells past their antioxidant capacity could be a selective therapeutic vulnerability. The findings reported here, in which mefloquine depleted glutathione, raised intracellular iron roughly threefold, and increased the lipid peroxidation markers 4-hydroxynonenal and malondialdehyde approximately twofold, fit squarely within this framework and suggest that the drug acts upstream of the canonical ferroptosis execution machinery by damaging mitochondria.

The mitochondrial dimension of the mechanism deserves particular emphasis. Mitochondria are both the principal source of ATP through oxidative phosphorylation and a major site of reactive oxygen species generation, since the electron transport chain leaks electrons that reduce oxygen to superoxide. The Seahorse metabolic flux measurements showing a 67 percent reduction in basal oxygen consumption rate and a 72 percent drop in ATP production indicate that mefloquine compromises the respiratory capacity of colorectal cancer cells at clinically relevant concentrations. When the electron transport chain is impaired, electrons accumulate at damaged complexes and are increasingly transferred to oxygen, generating superoxide that is converted to hydrogen peroxide and, in the presence of labile iron, to hydroxyl radicals. Mitochondrial reactive oxygen species rose from 22.9 percent to 47.4 percent of cells in the treated populations, a shift large enough to overwhelm antioxidant buffering and initiate the peroxidation of polyunsaturated fatty acids in mitochondrial and plasma membranes. Transmission electron microscopy in the study captured the characteristic ultrastructural signature of ferroptosis, including shrunken mitochondria with condensed membranes and reduced cristae, which distinguishes this death mode from the nuclear condensation typical of apoptosis.

The bioinformatics pipeline that led to the identification of mefloquine illustrates how modern integrative approaches can compress the early stages of drug discovery. By combining four Gene Expression Omnibus microarray datasets with RNA-sequencing data from 382 colorectal tumors in The Cancer Genome Atlas, the investigators constructed a consensus picture of transcriptional dysregulation that no single cohort could provide. Weighted gene co-expression network analysis groups genes into modules whose correlated expression patterns often reflect shared biological functions, while support vector machine recursive feature elimination iteratively removes features with the least discriminative power, yielding a compact set of genes that separate tumor from normal tissue. Cox proportional hazards regression and Lasso regularization then narrowed the candidates to those with prognostic value, producing twenty marker genes partitioned into occurrence-related, development-related, and prognosis-related categories. Feeding the corresponding upregulated gene signatures into the Connectivity Map, a reference database that matches expression profiles against those induced by thousands of small molecules, reversed the disease signature and nominated mefloquine among ten candidate repurposable drugs. This reversal logic, sometimes called signature reversion, has a strong track record in oncology because compounds that transcriptionally push a diseased state back toward normality frequently carry therapeutic activity.

The breadth of the in vitro validation strengthens the case that the effect is not an artifact of a single cell line. An IC50 below 10 micromolar across all seven colorectal cancer lines tested indicates broad-spectrum activity that is unlikely to depend on one specific genetic background. Functional assays confirmed that mefloquine-treated cells lost proliferative capacity, showed reduced migration in wound healing and Transwell experiments, and died through a process consistent with ferroptosis rather than apoptosis or necrosis. The in vivo arm of the study is particularly noteworthy because oral dosing was used, which matters for a repurposing strategy since mefloquine is already formulated for oral administration in malaria treatment. A 54.5 percent reduction in xenograft tumor weight at 50 milligrams per kilogram with no appreciable toxicity suggests a therapeutic window that could be explored clinically, although xenograft models in immunocompromised mice cannot fully reproduce the tumor microenvironment, immune interactions, and pharmacokinetics of human disease.

Several mechanistic details merit closer attention from readers evaluating the translational potential of these findings. The interplay between iron and glutathione is central to ferroptosis regulation. Intracellular iron rose approximately threefold after mefloquine treatment, which would amplify Fenton chemistry and increase the flux of hydroxyl radicals capable of abstracting hydrogen atoms from membrane lipids. Simultaneously, glutathione depletion removes the primary substrate for glutathione peroxidase 4, disabling the main enzymatic brake on lipid peroxidation. The observed upregulation of heme oxygenase 1 and the glutamate cysteine ligase catalytic subunit, both of which are associated with oxidative stress responses and glutathione biosynthesis, can be interpreted as a compensatory attempt by the cells to restore redox homeostasis, and their induction actually serves as a molecular fingerprint of ferroptotic stress. The accumulation of 4-hydroxynonenal, a reactive aldehyde generated during peroxidation of omega-6 polyunsaturated fatty acids, is particularly damaging because it forms covalent adducts with proteins and nucleic acids, propagating injury beyond the lipid compartment where it originates.

From a clinical standpoint, the fact that mefloquine is already approved for malaria prophylaxis and treatment in many countries means that its human pharmacokinetics, dosing ranges, and adverse event profile are documented, which could accelerate early-phase trials in colorectal cancer. However, several caveats should temper expectations. The doses used in xenograft studies often exceed those tolerated during chronic human use, and mefloquine is known to cause neuropsychiatric side effects at antimalarial doses, including dizziness, sleep disturbance, and rarely more severe reactions, which could limit the doses achievable in oncology patients. Furthermore, the pharmacokinetic behavior of mefloquine, with its long half-life and extensive tissue binding, may differ substantially between malaria parasitemia and solid tumor contexts. Whether plasma concentrations achievable in humans can reproduce the sub-10 micromolar cytotoxicity observed in vitro remains an open question that only carefully designed phase 1 studies can answer.

The marker genes identified through the bioinformatics pipeline may also have independent value beyond drug selection. Occurrence-related genes could serve as early detection biomarkers, development-related genes might track disease progression, and prognosis-related genes could stratify patients by risk. Receiver operating characteristic analysis in the study demonstrated diagnostic discrimination, and validation in independent cohorts would be the logical next step before any clinical deployment. It is also worth noting that colorectal cancer is molecularly heterogeneous, encompassing microsatellite-stable and microsatellite-instable subtypes with different immune landscapes and treatment responses, so the ferroptosis-inducing effect of mefloquine may not apply uniformly across all molecular subgroups. Patients with deficient mismatch repair tumors, for example, often respond well to immune checkpoint inhibitors and might not need alternative cytotoxic strategies, whereas microsatellite-stable tumors with limited immunotherapy options could represent the population most likely to benefit from a ferroptosis-inducing agent.

Looking ahead, the convergence of computational drug repositioning, mechanistic validation of cell death pathways, and preclinical efficacy testing exemplified by this study reflects a broader trend in oncology research. As public datasets continue to grow and machine learning methods become more sophisticated, the bottleneck in drug repurposing is shifting from target identification to rigorous biological validation and clinical translation. The demonstration that a well-known antimalarial can trigger ferroptosis through mitochondrial damage adds a new entry to the expanding list of non-cancer drugs with unexpected anticancer properties and provides a mechanistic rationale that could guide combination strategies, for instance pairing mefloquine with agents that further deplete glutathione or that enhance iron accumulation in tumor cells. Such rational combinations, grounded in a clear understanding of the ferroptotic machinery, may ultimately prove more effective than single-agent approaches and could open new therapeutic avenues for patients with colorectal cancer whose tumors have progressed beyond current standards of care.

Subject of Research: Drug repurposing and ferroptosis induction in colorectal cancer

Article Title: Integrated bioinformatics and experimental validation identifies the antimalarial drug mefloquine as a ferroptosis inducer in colorectal cancer via mitochondrial damage and oxidative stress-mediated lipid peroxidation

Article References: Integrated bioinformatics and experimental validation identifies the antimalarial drug mefloquine as a ferroptosis inducer in colorectal cancer via mitochondrial damage and oxidative stress-mediated lipid peroxidation. (n.d.). https://doi.org/10.1186/s12885-026-17144-2

Image Credits: AI Generated

DOI: 10.1186/s12885-026-17144-2

Keywords: Integrated, bioinformatics, experimental, validation, identifies, antimalarial, drug, mefloquine, ferroptosis, inducer, colorectal, cancer