The team began by creating laboratory models that faithfully mimic what happens in patients undergoing fractionated radiotherapy. They exposed two head and neck cancer cell lines, OECM1 and Detroit, to repeated rounds of radiation over an extended period, generating isogenic radioresistant sublines. This long-term fractionated irradiation approach is important because clinical radiotherapy is delivered in many small doses over weeks, and acute single-dose experiments can miss the adaptive changes that accumulate during a real treatment course. With these paired sensitive and resistant cell lines in hand, the researchers performed global microRNA profiling to determine which of these small regulatory RNAs change consistently when cells become radioresistant.

MicroRNAs are short, non-coding RNA molecules that regulate gene expression after transcription by binding target messenger RNAs and suppressing their translation or promoting their degradation. A single microRNA can regulate dozens or even hundreds of target genes, which makes them powerful coordinators of cellular behavior. The profiling revealed a remarkably consistent signature: 25 microRNAs were altered in the radioresistant sublines of both cell lines. Twelve were upregulated and behaved as oncogenic microRNAs, or oncomiRs, while thirteen were downregulated and appeared to function as tumor-suppressive microRNAs. The fact that the same signature emerged in two independent cell lines suggested that these changes are not random artifacts but a coordinated program underlying radioresistance.

To understand what this program actually does, the researchers performed integrative target prediction, pathway enrichment analysis, and network construction, linking each microRNA to its predicted messenger RNA targets. The analysis showed that the 25 altered microRNAs converge on a set of oncogenic modules with clear relevance to aggressive cancer biology: receptor tyrosine kinase signaling, cell motility, and stress response and cancer stemness pathways. At the center of the network sat familiar villains of head and neck cancer, including the epidermal growth factor receptor EGFR, the insulin-like growth factor 1 receptor IGF1R, and the transcriptional regulator MYC. The refined network distilled down to 68 microRNA–messenger RNA interaction pairs, with individual microRNAs such as miR-199b-5p and miR-522-3p emerging as key regulatory nodes.

Crucially, the network was not merely a computational exercise. When the researchers ectopically overexpressed the key regulatory microRNAs in head and neck cancer cells, the cells became significantly more sensitive to radiation. This functional validation demonstrated that the microRNA nodes identified through bioinformatics genuinely control radiosensitivity, strengthening the case that the network represents a real biological mechanism rather than a statistical pattern. It also raised an obvious therapeutic question: if a network of microRNAs and their targets maintains radioresistance, could a drug be found that collapses the entire network at once?

Answering that question is where the study takes a particularly modern turn. Rather than screening thousands of compounds blindly, the team turned to transcriptome-guided drug repurposing using the Connectivity Map platform. This approach inverts the traditional drug discovery logic: instead of asking which chemicals kill cancer cells, it asks which approved or investigational drugs produce a gene expression signature opposite to the disease signature of interest. When the radioresistant microRNA–messenger RNA network signature was queried against the drug database, one compound rose decisively to the top: I-OMe-AG-538, an inhibitor of IGF1R, with a strong negative connectivity score of tau equal to minus 87, indicating that the drug’s transcriptional effects run counter to the resistance program.

The choice of IGF1R as the target makes biological sense in light of the network analysis. The receptor sat among the central hubs of the resistance network, and blocking it would be expected to weaken multiple oncogenic modules simultaneously. Experimental validation confirmed this prediction. Treating radioresistant cells with I-OMe-AG-538 dose-dependently suppressed the phosphorylation of IGF1R and its downstream signaling partner Erk, cutting off the growth and survival signals that the receptor transmits. More strikingly, the drug reprogrammed the microRNA landscape itself: oncomiRs that had been elevated in resistant cells were pushed down, while tumor-suppressive microRNAs that had been lost were restored. In effect, the compound nudged the entire regulatory network back toward a radiation-sensitive state.

The drug also attacked a second pillar of radioresistance: redox homeostasis. Ionizing radiation kills cells largely by generating reactive oxygen species, and resistant tumors often survive by scavenging these molecules. I-OMe-AG-538 elevated intracellular reactive oxygen species levels in the treated cells, overwhelming one of the defenses that resistant tumors rely on. In clonogenic assays—the gold-standard test of a cell’s ability to survive radiation and form new colonies—the combination of the drug with irradiation synergistically increased radiosensitivity, meaning the two treatments worked together more powerfully than either alone. Supporting the clinical relevance of these findings, an analysis of The Cancer Genome Atlas head and neck squamous cell carcinoma dataset showed that patients whose tumors expressed high levels of IGF1R had poorer prognosis, and that high IGF1R expression correlated with upregulation of genes that scavenge reactive oxygen species.

The broader significance of the work lies in its strategy. Drug repurposing offers a faster and cheaper route to the clinic than de novo drug development, because repurposed compounds often have established safety profiles. By anchoring the search in a mechanistically defined microRNA–messenger RNA network rather than a generic gene expression profile, the researchers gained both a drug candidate and a biological rationale for why it should work. The study also underscores how microRNAs can serve as both biomarkers and functional mediators of treatment resistance, and how targeting upstream signaling hubs like IGF1R can produce downstream effects across an entire regulatory network, including the microRNA layer itself.

Considerable work remains before patients could benefit. The findings come from cell line models and computational analyses, and the drug’s ability to sensitize tumors in living animals and in combination with standard chemoradiotherapy regimens will need to be demonstrated. Dosing, toxicity, and potential interactions with other receptor tyrosine kinase inhibitors already used in head and neck cancer all require study. Nevertheless, the research provides a coherent roadmap: define the resistance network, validate its key nodes, find a compound that reverses the network’s signature, and confirm that the compound restores radiation sensitivity while disrupting both oncogenic signaling and antioxidant defenses. For patients whose head and neck cancers refuse to yield to radiation, that roadmap offers a genuinely new direction—one in which an old drug, guided by a new map, may help radiation finish the job it was meant to do.

Subject of Research: miRNA–mRNA regulatory network underlying radioresistance in head and neck cancer and transcriptome-guided identification of the IGF1R inhibitor I-OMe-AG-538 as a radiosensitizer

Article Title: Targeting a miRNA–mRNA regulatory network to overcome radioresistance in head and neck cancer: identification of I-OMe-AG-538 via transcriptome-guided drug repurposing

Article References: You, G.-R., Chang, J. T., Huang, H.-H., Shen, E. Y.-L., Chen, Y.-J., Li, Y.-L., & Cheng, A.-J. (2026). Targeting a miRNA–mRNA regulatory network to overcome radioresistance in head and neck cancer: identification of I-OMe-AG-538 via transcriptome-guided drug repurposing. Cell Death Discovery. https://doi.org/10.1038/s41420-026-03357-8

Image Credits: AI Generated

DOI: 10.1038/s41420-026-03357-8

Keywords: head and neck cancer, radioresistance, microRNA, IGF1R, I-OMe-AG-538, drug repurposing, Connectivity Map, radiosensitizer, reactive oxygen species, EGFR, transcriptomics, radiotherapy