Drug-tolerant persister cells are a recognized barrier to durable responses in patients with EGFR-mutant lung adenocarcinoma treated with tyrosine kinase inhibitors (TKIs). These persister cells survive initial targeted therapy and can contribute to relapse or acquired resistance. The study aimed to define the molecular features and actionable vulnerabilities of the persister cell state induced by targeted therapy.
The authors combined public transcriptomic datasets derived from tumour cell-based models treated with EGFR inhibitors or other targeted agents acting within the RTK/RAS/MAPK pathway. They applied machine-learning–based feature selection together with survival analysis and pharmacogenomic screening using the Cancer Dependency Map. This integrated strategy was used to identify gene-level signatures associated with therapy‑induced persisters and to nominate candidate drugs that could selectively impair persister viability.
From the integrated transcriptomic and computational analyses, a four-gene signature associated with the targeted therapy–induced persister state was defined. The four genes are: B-cell translocation gene 1 protein, inhibitor of growth protein 4, proline-rich nuclear receptor coactivator 1, and calcium-binding and coiled-coil domain-containing protein 1. This signature was reported to mark a persister phenotype across the analysed cell-model datasets.
The study linked the four-gene persister-associated signature to poorer patient survival through survival analysis of available datasets. The presence of the signature was associated with worse outcome, supporting clinical relevance of the persister transcriptional program. Specific dataset identifiers, cohort sizes and statistical measures were not provided in the abstract and therefore are not reported here.
Experimental validation focused on two EGFR‑mutant lung adenocarcinoma cell lines, PC‑9 and HCC827. Using reverse transcription‑quantitative PCR, the authors showed rapid induction of the four signature genes during treatment with osimertinib, a clinically used third‑generation EGFR TKI. These observations support the computational finding that these genes are upregulated as part of a TKI-induced persister program.
To test functional contribution of each gene to persister formation, the investigators employed doxycycline‑inducible short hairpin RNA (shRNA) knockdown constructs for individual genes. Depletion of each signature gene markedly impaired formation of drug-tolerant persister cells in the tested models. Additionally, β‑galactosidase staining assays were used as part of the functional assessment; the abstract reports that knockdown reduced persister features, but detailed quantitative readouts are not provided in the source abstract.
Pharmacogenomic analysis leveraging the Cancer Dependency Map nominated ZLN005 as a candidate metabolic modulator that could synergize with osimertinib to reduce persister cell viability. ZLN005 is described as a peroxisome proliferator‑activated receptor γ coactivator‑1α (PGC‑1α) agonist. The authors validated the combinatorial effect experimentally in the PC‑9 and HCC827 cell models, where co-treatment with ZLN005 and osimertinib reduced persister viability compared with single agents.
Mechanistically, ZLN005 was found to enhance mitochondrial oxidative metabolism and increase reactive oxygen species (ROS) accumulation. Elevated ROS correlated with activation of caspase‑3 and induction of gasdermin E (GSDME)-associated pyroptotic cell death. The sequential events reported in the abstract are: ZLN005 drives mitochondrial oxidative metabolism → increases ROS → activates caspase‑3 → triggers GSDME-mediated pyroptosis, resulting in loss of persister cell viability. The study therefore links metabolic modulation to a programmed lytic death pathway as an exploitable vulnerability in TKI-induced persisters.
Collectively, the findings define both transcriptional and metabolic characteristics of targeted therapy–induced persister cells in EGFR‑mutant lung adenocarcinoma and identify a metabolic vulnerability to ZLN005. The work supports the concept that combining EGFR TKIs such as osimertinib with metabolic modulators that enhance mitochondrial oxidation and ROS can activate caspase‑3/GSDME-dependent pyroptosis and thereby reduce persister survival. This points toward a potential therapeutic strategy to limit persister-mediated relapse.
The abstract does not report detailed information on dataset identifiers, sample sizes, quantitative effect sizes, in vivo validation, safety or translational clinical data. These details are necessary to fully assess robustness, reproducibility and potential clinical applicability but were not available in the source abstract provided.