Background:
Resistance to tyrosine kinase inhibitors (TKIs) in EGFR-mutant lung cancer remains a major challenge. Osimertinib, a third-generation EGFR-TKI, improves outcomes for patients progressing on earlier TKIs, targeting both sensitizing mutations and the T790M resistance mutation. However, acquired resistance to Osimertinib eventually develops. While many studies have investigated mechanisms of Osimertinib resistance in the PC9 cell line, few have explored this at the single-cell level. Single-cell analysis can reveal cancer cell heterogeneity, which drives varying responses to TKIs, and helps identify therapeutic targets for more personalized treatment strategies.
 
Objectives:
This study aims to identify and characterize heterogeneous responses in the PC9 cell line following treatment with the EGFR-TKI - Osimertinib.
 
Methods:
Single-cell RNA-sequencing (scRNA-seq) from PC9 lung cancer cell line and DFCI282 PDX lung cancer model, before and after Osimertinib treatment, were obtained from the Gene Expression Omnibus (GSE302284). Following preprocessing and quality control, Uniform Manifold Approximation and Projection (UMAP) for dimensionality reduction and hierarchical clustering were applied to identify distinct cellular subpopulations.
Differentially expressed genes (DEGs) were identified using the findMarkers function from the scran package. For each treated cluster, up-regulated DEGs compared to the other two clusters were analyzed for Gene Ontology Biological Process (GO:BP) terms using Gene Set Enrichment Analysis (GSEA). DEGs from treated clusters, compared to control group, were classified into up-regulated and down-regulated categories based on logFC. From each catergory, common DEGs across all clusters and unique DEGs for each cluster were identified, after filtering with p-value < 0.05 and |logFC| > 0.25 to avoid identical DEGs across clusters due to the shared original dataset. The enrichGO function was used to identify enriched pathways for these DEGs. Both pathway analyses were performed using the clusterProfiler package and org.Hs.eg.db annotation database.
 
Results:
To address the challenge of Osimertinib resistance in EGFR-mutant lung cancer, this study utilized single-cell analysis to characterize the heterogeneous responses of the PC9 cell line. The analysis revealed distinct cellular behaviors: a drug-sensitive apoptotic subpopulation exhibiting downregulated aerobic respiration and ATP synthesis consistent with cell death pathways, and a drug-tolerant proliferative subpopulation characterized by sustained protein synthesis and metabolic activity despite treatment. This functional divergence was further validated in the DFCI282 PDX model, confirming the existence of adaptive cell states. Furthermore, a gene regulatory network was constructed based on the differentially expressed genes between these subpopulations to identify key transcription factors driving the shift toward the adaptive phenotype. Targeting these master regulators holds clinical potential as a strategy to block the transcriptional programs responsible for drug tolerance, thereby preventing the emergence of acquired resistance.
 
Conclusion/Implications:
These findings highlight the heterogeneous response of PC9 lung cancer cell to Osimertinib and emphasize the need for personalized treatment strategies based on distinct tumor cell behaviors.