Sorafenib-associated translation reprogramming in hepatocellular carcinoma cells

Abstract

Sorafenib (Sfb) is a multikinase inhibitor regularly used for the management of patients with advanced hepatocellular carcinoma (HCC) that has been shown to increase very modestly life expectancy. We have shown that Sfb inhibits protein synthesis at the level of initiation in cancer cells. However, the global snapshot of mRNA translation following Sorafenib-treatment has not been explored so far. In this study, we performed a genome-wide polysome profiling analysis in Sfb-treated HCC cells and demonstrated that, despite global translation repression, a set of different genes remain efficiently translated or are even translationally induced. We reveal that, in response to Sfb inhibition, translation is tuned, which strongly correlates with the presence of established mRNA cis-acting elements and the corresponding protein factors that recognize them, including DAP5 and ARE-binding proteins. At the level of biological processes, Sfb leads to the translational down-regulation of key cellular activities, such as those related to the mitochondrial metabolism and the collagen synthesis, and the translational up-regulation of pathways associated with the adaptation and survival of cells in response to the Sfb-induced stress. Our findings indicate that Sfb induces an adaptive reprogramming of translation and provides valuable information that can facilitate the analysis of other drugs for the development of novel combined treatment strategies based on Sfb therapy.

Keywords: Hepatocellular carcinoma; Sorafenib; m6A methylation; polysome profiling; translation.

Figure 1.

Ribo-Seq analysis of untreated and Sfb-treated HCC cells. A. Scheme of the workflow for the polysome profile analysis and calculation of the translation level (TL) parameter. B. Dot plots showing the TL of control or untreated (X-axis) and Sfb-treated samples (Y-axis). Genes whose translation is significantly influenced in a positive or a negative manner in each single or in combinations condition are highlighted. C. Proportional Venn diagram showing the number of genes whose translation is positively or negatively influenced in each condition (p- value < 0.05; TL < − 0.5 or > 0.5).

Figure 2.

Functional enrichment analysis. Bubble plots depicting the over-representation analysis (ORA) of different mRNA features or cis-acting elements (from left to right, uORF containing mRnas, CPEB4 target mRnas, are-containing mRNAs in their 3’ UTRs, DAP5 target mRNAs and IRES containing mRNAs) in the gene categories defined in Figure 1B. In each plot, the -log₁₀ (p-value) of the ORA is shown in the X-axis. A colour scale is applied. The size of each bubble corresponds to the enrichment of each feature as the number of genes with the studied feature in each category versus the total number of genes with the feature and normalized by the number of genes in the group versus the total number of genes represented in the RNA-Seq file.

Figure 3.

Definition of Sfb hypersensitive genes. A. Scatter plot between the TL in Sfb-treated samples and the log₁₀ of the CDS length. Note that a population of genes with low negative TL, named Sfb hypersensitive genes (in blue), does not follow the general tendency. B. Scatter plot between the difference in TL of the Sfb-treated versus the control samples and the log₁₀ of the CDS length. Sfb hypersensitive genes are again highlighted in blue.

Figure 4.

Abundance of non-m⁶A methylated genes in the different categories of our analysis. Left panel, bubble plot showing the ORA of non-m⁶A methylated mRNAs within the different gene categories defined in Figure 1B and the group of genes hypersensitive to Sfb. Right panel, scatter plot between the TL in Sfb-treated samples and the log₁₀ of the CDS length. Genes coding non-m [6]A methylated mRNAs are highlighted in red.

Figure 5.

REACTOME functional annotation for genes with an opposite direction identified in our analysis. Bubble plots showing the ORA of the top REACTOME categories in the genes with an opposite direction in Sfb-treated and untreated control samples. Left panel, Sfb-up and control-down category. Right panel, Sfb-down and control-up category.

Figure 6.

REACTOME functional annotation for specific genes identified in our analysis. Bubble plot showing the ORA of the top REACTOME categories in the genes with an either up-regulated (left panel) and down-regulated (right panel) translation in the Sfb-treated samples but with a not up-regulated (left panel) and down-regulated (right panel) translation in control samples, respectively.