Hepatic diseases at the molecular level : identifying novel targets and biomarkers through transcriptomics

Abstract

In the last few decades, the prevalence of obesity and metabolic diseases have dramatically increased and pose a major threat to global healthcare systems. The chronic excess of carbohydrates and fat is a risk factor for many diseases, including cardiovascular disease, several types of cancer, and metabolic dysfunction-associated steatotic liver disease (MASLD). Concomitant with the obesity pandemic, MASLD has become the most prevalent liver disease and affects more than a third of the global population. Early MASLD stages are characterized by hepatic lipid droplet accumulation that are considered asymptomatic, however, more severe stages include liver inflammation, scarring, liver cirrhosis and even hepatocellular carcinoma. Apart from recently approved drugs like Wegovy, weight loss is the only effective treatment, hence creating a demand for new therapeutic targets for MASLD as well as reliable diagnostic biomarkers.

In the work presented in this doctoral thesis, we explored novel potential therapeutic targets and diagnostic biomarkers in MASLD and hepatocellular carcinoma by combining molecular biology techniques, high-throughput sequencing-based methods such as transcriptomics, and computational analyses.

In paper 1, we took advantage of the observation that women of fertile age are less frequently diagnosed with MASLD than women after menopause or men, and investigated the protective effects of estrogen signaling in MASLD. We found that only male mice fed a high-fat diet (HFD) developed intrahepatic lipid droplets, and treatment with different estrogen-like molecules ameliorated this phenotype. Using transcriptomics, we dissected the altered hepatic gene and pathway landscape upon activated estrogen signaling and determined affected cell types by integrating publicly available single cell RNA-seq atlases. Since transcriptomic changes can occur via modulation of the epigenomic landscape, we profiled the histone modifications histone 3 lysine 27 acetylation (H3K27ac) and histone 3 lysine 4 trimethylation (H3K4me3) using chromatin immunoprecipitation followed by sequencing (ChIP-seq). We found that estrogen treatment also partly reversed diet-induced changes of these histone modifications. Additionally, we identified enhancer gene pairs that were sensitive to diet and estrogen signaling. In a publicly available RNA-seq dataset comprising 206 patients with MASLD, we confirmed that many genes altered in mice followed similar gene expression trends in human disease. Lastly, we found that the TEA associated domain 1 (TEAD1) gene, encoding an important transcription factor of the Hippo pathway, was among the identified candidate genes. TEAD1 inhibition in a mouse cell line and primary human hepatocyte-derived spheroids demonstrated that TEAD1 regulates hepatic steatosis and could be a promising therapeutic target.

In paper 2, we uncovered novel serological diagnostic biomarkers for MASLD using a transcriptomics approach. We re-analyzed two RNA-seq datasets comprising more than 300 patients with MASLD, and identified genes that encoded for secreted proteins and were upregulated in disease compared to healthy individuals. We found that the protein levels of our top-ranked candidate, Sulfatase 2 (SULF2), were significantly increased in the serum of 31 patients with MASLD compared to 11 healthy individuals. These results suggest that SULF2 is a promising candidate to non-invasively diagnose MASLD.

In paper 3, we investigated the roles of RNA binding proteins (RBPs) in hepatocellular carcinoma (HCC) since this class of proteins has been implicated to regulate central cell functions. After shortlisting candidate RBPs that were associated with low patient survival, we depleted the mRNAs of ten RBPs in HCC cell models and characterized the effect on cell survival and the transcriptome. A noncanonical RBP, i.e., RBP without harboring known RNA-binding domains (RBDs), called T-complex protein 1 subunit gamma (CCT3) exhibited the strongest effect on proliferation and surprisingly deregulated many noncoding (nc)RNAs. One long noncoding (lnc)RNA called LINC00326 was upregulated upon CCT3 depletion, and we hypothesized that this link could be responsible for cancer cell death. We mimicked the LINC00326 upregulation by using a CRISPR activation (CRISPRa) approach that endogenously increased LINC00326 RNA levels and found a strong detrimental effect on HCC cell survival. We found evidence that CCT3 and LINC00326 physically interact. Moreover, we determined that CCT3 and LINC00326 were functionally dependent. The simultaneous knockdown of CCT3 and LINC00326 partially rescued the phenotype on cell survival, indicating that CCT3 killed HCC cells by upregulation of LINC00326. Lastly, we investigated through which mechanism the CCT3-LINC00326 axis functioned and uncovered higher lipid peroxidation levels and reduced intracellular storage lipids upon CCT3 knockdown or LINC00326 overexpression, suggesting that this favored HCC cell death.

In paper 4, we developed a cost-effective and easy-to-apply method to overcome low transfection efficiencies in experiments dependent on large plasmids. We found out that the addition of small plasmids to large CRISPR-Cas9 plasmids increased transfection efficiency and reduced cell death. We uncovered that the method elevated transfection efficiency in all tested cell lines and types, including several hard-to-transfect cell lines and primary cells. The observed benefit was independent of gene content on the small plasmid.