Fig. 3: ECM stiffness regulates G6PD expression.

A Venn diagram illustrating the overlap between genes related to ECM stiffness and those associated with sorafenib resistance. B Analysis of the top 20 KEGG pathways for genes related to ECM stiffness and sorafenib resistance. C Cross-validation plot for penalty terms. The cross-validation was used to select the optimum value of the regularization parameter (λ) in the LASSO regression model. D LASSO coefficient profiles of genes associated with ECM stiffness and sorafenib resistance. A coefficient profile plot was produced against the log(λ) sequence. A vertical line was drawn at the selected optimal parameter (λ), which resulted in 12 nonzero coefficients. E A forest plot of 12 genes associated with ECM stiffness and sorafenib resistance. F Coefficients of these 12 genes. G Real-time qPCR analysis of gene expression related to ECM stiffness under soft (0.5 kPa) or stiff (12 kPa) matrix conditions. H Western blotting analysis of G6PD and AKR1B20 expression under soft (0.5 kPa) or stiff (12 kPa) matrix conditions. β-actin was used as a loading control and quantification data were shown on the right panel. I Representative images of H&E, Collagen I, α-SMA, Masson’s trichrome, Picrosirius red, AKR1B10, and G6PD staining in HCC tissues from patients who exhibited low stiffness and responsiveness to sorafenib and those with high stiffness and resistance to sorafenib. Scale bar, 200 and 500 μm as indicated. J Quantification of collagen I, α-SMA, Masson’s trichrome, and Picrosirius red area between 10 sorafenib-resistant samples with high ECM and 10 sorafenib-sensitive samples. K Association between AKR1B10 and G6PD expression levels and the degree of ECM stiffness in HCC tissue. ns not significance; *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001.