Changes in genes can be observed at early stages in drug-treated sensitive cells, which aids in our understanding and prevention of the development of drug resistance. Previous reports have used Time-resolved dual RNA-seq to study the impact of Streptococcus pneumoniae on lung epithelium during early infection, and related genes in epithelial cells also show differential expression as the infection persists[40]. The prospect of studying early activity changes in cells treated in a certain way using sequencing technology is enticing, and as sequencing technology advances, research in this area will become increasingly in-depth.
In this study, we analyzed gene changes in sorafenib-treated hepatocellular carcinoma cells (SMMC-7721) in the two hours using KAS-seq and plotted KAS-seq profiles of gene coding regions under different treatment times. We found that the signal abundance decreased from no drug treatment to one hour of treatment, and then from one hour to two hours of treatment, reflecting that the cell’s response to changes in the external environment after drug treatment varies with different treatment times.
Subsequently, based on the KAS-seq signal distribution (Fig. 1B), we selected the time point (1h) corresponding to the highest signal intensity for differential enrichment analysis with the control group. In the KEGG enrichment analysis, differentially expressed genes were enriched in pathways closely related to tumors, such as pathways in cancer, MAPK signaling pathway, cAMP signaling pathway, and tight junction, confirming the feasibility of KAS-seq. Furthermore, through the molecular function (MF), we found that most of the up- and down-regulated differentially expressed genes were concentrated in protein binding. This implies that further research may be needed on the protein interactions of these genes and how these interactions affect sorafenib resistance.
Clinical samples, which come directly from patients and are typically associated with specific diseases or pathological processes, can more accurately reflect the conditions within a biological organism. Given that the effectiveness of chemotherapy is primarily limited by drug resistance, studying patients for whom drug treatment is ineffective or effective can help us understand the resistance mechanism of sorafenib[41]. Therefore, we selected patients from the clinical sample dataset (GSE109211) who were ineffective or effective in sorafenib treatment for comparison, and obtained differentially expressed ERGs. We performed enrichment analysis on ERGs and found that upregulated genes in KEGG were mainly concentrated in metabolic pathways, while downregulated genes were concentrated in olfactory transduction. Notably, in MF, upregulated differentially expressed genes were mainly enriched in protein binding, while downregulated differentially expressed genes were mainly related to G-protein coupled receptor activity.
Through survival analysis, we found that the expression levels of 7 genes (ACTB, CFL1, ACTG1, ACTN1, WDR1, TAGLN2, HSPA8) were significantly associated with poor prognosis in patients. High gene expression often leads to worse patient outcomes. These genes were identified through differential expression analysis between patients who were responsive and non-responsive to Sorafenib treatment. This suggests that genes potentially related to Sorafenib resistance may drive disease progression, further indicating a close relationship between the differentially expressed genes we analyzed and Sorafenib resistance. We then validated these findings at both the transcriptional and protein levels.
To further understand the drug resistance mechanisms involved with 7 hub ERGs, we conducted analyses using GeneMANIA, GO, and KEGG, and found that these genes are closely related to the cytoskeleton. According to previous reports, these genes promote the invasiveness and metastasis of tumors, which is often related to changes in the cytoskeleton[42–48]. Previous studies have reported that drug treatment stress can cause molecular and phenotypic changes in tumor cells, i.e., cellular plasticity, thereby inducing tumor drug resistance[24]. This suggests that the remodeling of the cytoskeleton may be related to sorafenib resistance in hepatocellular carcinoma.
Furthermore, when we performed GSEA analysis on these cytoskeleton-related genes using patient samples treated with sorafenib from the GSE109211 dataset, we found that many genes were primarily enriched in ubiquitin-mediated proteolysis. This may suggest their important role in the cytoskeleton.