Evolutionary trajectory of 5-ethynyl-2’-deoxyuridine resistance in HeLa cells reveals a non-linear transition from acute replication stress to transcriptomic desensitization

Resistance to nucleoside-like compounds in cancer treatment is a significant problem. However, the transcriptional pathways by which cancer cells adapt under increasing drug stress are not yet sufficiently elucidated for many clinically promising molecules. 5-Ethynyl-2′-deoxyuridine (EdU) is a thymidine analogue that exhibits a direct antitumor effect and can cross the blood-brain barrier. This compound leads to cell death by inhibiting the thymidylate synthase enzyme, arresting the replication fork, and initiating a futile cycle caused by nucleotide excision repair. To fully map the resistance process to EdU at a single concentration level, adaptive laboratory evolution (ALE) was applied to HeLa cells, and six different resistant cell line stages were obtained with gradually increasing doses of resistance from 0.1 µM to 80 µM EdU. Resistance was characterized by: cell viability tests, quantification of EdU incorporation into genomic DNA by immune-slot blot, morphological examination by bright-field microscopy, and RNA sequencing of all lines under both untreated and acute EdU application (10 µM, 24 h). In this study, five resistance phases differing in mechanism were identified: stress readiness, persistent cell-like AP-1/EGR arrest, AP-1 signaling re-stimulation and increased complexity, the phase where the integrated stress response led by ATF4 reaches its peak, and finally, transcriptional silence in EdU response. Remarkably, cells that have acquired almost complete resistance to EdU-induced cell death do not develop any canonical AP-1, inflammatory, or DNA damage-related transcriptional responses after acute drug administration; we define this as transcriptomic desensitization. This state differs entirely in its mechanism from conventional nucleoside analog resistance: there is no upregulation in either thymidylate synthase or genes involved in DNA repair. Instead, the resistance mechanism relies on two key elements: metabolic exclusion resulting from the suppression of the SLC22A17 transporter and the stepwise remodeling of cell identity towards a partial epithelial-mesenchymal transition (EMT) phenotype. These findings make EdU resistance a useful and investigable model for studying transcriptional adaptation processes to genotoxic nucleoside analogs, particularly those directly associated with glioblastoma combination therapies.