Fig. 7
The abundance of QSOX1 controls sensitivity of NSCLCs to gefitinib. a, b Representative images of QSOX1 IHC staining (a) and the quantitative H-scores (b) in pre-treatment and post-treatment biopsies of NSCLC patients with intrinsic resistance to gefitinib. Image magnification: 200 × (upper panel) and 400 × (lower panel). c Sensitive and resistant cells were treated with gefitinib (1 µM) or osimertinib (1 µM) for 24 h and QSOX1 protein level was detected by western blotting. d H1650 and H1975 cells transfected with siRNA targeting QSOX1 (siQSOX1) or non-targeting control (siCtrl) were exposed to gefitinib (1 µM) for 72 h and cell viability was detected by CCK-8 assay. e, f H1650 cells pre-treated with shNSUN2 were stably transfected with wild type QSOX1 (QSOX1-WT) or QSOX1 with the mutated m5C sites (QSOX1-Mut) and cell proliferation was detected by CCK-8 (e) or colony formation assay (f). g, h Tumor growth curve (g) and tumor weights (h) of xenografts subcutaneously implanted with H1975-shCtrl and H1975-shQSOX1 cells in BALB/c nude mice (n = 7 per group). i Kaplan-Meier analysis of TCGA-LUAD dataset showing the combined high expression of NSUN2, YBX1 and QSOX1 predicting a poorer overall survival. P value (p = 0.016) was determined using a log-rank test. j Working model for aberrant m5C hypermethylation conferring intrinsic resistance to gefitinib in NSCLC via NSUN2/YBX1/QSOX1 axis. The results are shown as means ± SD. For c-e, n = 3 biological independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001.