BEHIND THE PAPER
N1-methyladenosinemethylation in tRNA drives liver tumourigenesis by regulating cholesterol metabolism
Yanying Wang, Zusen Fan
Institute of Biophysics, Chinese Academy of Sciences, China
Liver cancer is the fourth most common cancer-related death and is the sixth in terms of cancer morbidity worldwide1. It remains difficult to treat it because of lacking drugs for critical targets. Hepatocellular carcinoma (HCC), a cause of chronic infection of hepatitis B or C virus (HBV or HCV) or alcohol abuse, accounts for the majority of primary liver cancers. HCC is characterized by high recurrence and heterogeneity that is mainly caused by hierarchical organizations of tumour cells with cancer stem cells (CSCs)2. However, how liver CSCs maintain their self-renewal remains largely unclear. Our group has been worked on the exploration of self-renewal mechanisms of liver CSCs over ten years. We defined CD13+ and CD133+ cell subpopulation as a subset of liver CSCs. For the first time, we identified deregulated lncRNAs in liver CSCs and revealed their distinct roles in the self-renewal regulation of liver CSCs 3.
RNA modifications have recently become critical posttranscriptional regulators in tumour development. N1-methyladenosine (m1A) methylation is an essential and highly conserved RNA modification. m1A methylation can be catalyzed by a methyltransferase complex, containing TRMT6 and TRMT61A components4,5. We wanted to know whether m1A methylations played some roles in the self-renewal of liver CSCs and tumourigenesis. First, we analyzed the expression of TRMT6 and TRMT61A in different online datasets of HCC patient samples. We found that TRMT6 and TRMT61A levels were remarkably increased in tumour tissues compared to peri-tumour tissues. In TCGA dataset, liver cancer samples with highest levels of TRMT6 and TRMT61A in tumours exhibited advanced stages and worse survivals. Consistently, we confirmed their higher expression of mRNA and proteins in tumour tissues than in peri-tumour tissues based on our lab’s HCC samples. Then, using HCC tumour tissues and paired peri-tumour tissues from a cohort of HCC patients who had not received previous chemotherapy or radiotherapy, we found that m1A and TRMT6 levels were aberrantly elevated in HCC tumour tissues using immunohistochemical staining. Notably, m1A levels were much higher in poorly differentiated than in well differentiated HCC samples, especially most elevated in samples with microscopic vascular invasion. In addition, we found that liver CSCs (CD13+CD133+) displayed much higher m1A levels than non-CSCs (CD13-CD133-) from HCC samples and HCC cell lines via different analyses. Collectively, these results suggest that differentially expressed m1A in liver CSCs and non-CSCs may regulate the self-renewal of liver CSCs and liver tumorigenesis.
We then explored the role of m1A methylation in liver CSC self-renewal and tumorigenesis. To determine whether TRMT6/TRMT61A exerted function through its enzymatic activity, we generated their mutant constructs. We found that TRMT6 and TRMT61A depletion by shRNA-mediated knockdown dramatically reduced the self-renewal of liver CSCs and tumour growth of orthotopic xenografts. Notably, although forced expression of mutant TRMT6/TRMT61A could increase the protein levels of the complex, it could not rescue liver CSC self-renewal and tumour growth compared with wild type TRMT6/TRMT61A in TRMT6/TRMT61A depleted cells. We next generated Trmt6 and Trmt61a knockout mice using CRISPR/Cas9 approaches. We found that their knockout remarkably suppressed liver tumourigenesis in vivo. These data indicated that m1A positive liver CSCs are dependent on the methylation activity of TRMT6/TRMT61A complex. TRMT6/TRMT61A-mediated m1A methylation might serve as a target for HCC therapy.
Given that global m1A levels were elevated in liver CSCs, we next sought to interrogate differentially methylated tRNAs using an improved single-nucleotide resolution quantitative detection method (tRNA-m1A-seq). Interestingly, we found that tRNAs with high m1A58 methylation levels were similar between liver CSCs and non-CSCs. And for several tRNAs with medium levels of methylation, they had higher m1A58 levels in CSCs, but both global tRNA levels and individual tRNA level between CSCs and non-CSCs showed no significant difference. These data indicated that tRNA m1A modification but not tRNA expression level regulates liver CSCs. To further identify the transcripts whose translation and which pathway were regulated by these m1A58 elevated tRNAs, we performed ribosome profiling sequencing (Ribo-seq), genome-wide RNA-seq, quantitative proteomic analysis and lipidomic metabolite profiling analyses. We found PPARδ -cholesterol biosynthesis- Hedgehog signaling was key regulators of targeted tRNAs with differential m1A58 methylation in liver CSCs.
Our next question was whether TRMT6/TRMT61A complex mediated m1A modification could be a therapeutic target for HCC, which were of great interests for RNA modification, cancer, and pharmaceutical research fields. Based on our findings, we then screened potential inhibitors blocking TRMT6 and TRMT61A protein-protein interaction (PPI) from a FDA-approved drug bank, followed by subsequent screening using m1A methylation level as readout. We identified a small molecule compound thiram could efficiently impair m1A modifications and liver CSC self-renewal in vitro. Accordingly, it could reduce tumour growth and extend the survival time of tumor-bearing mice orthotopicly implanted with HCC patient-derived tumour cells (PDC). In addition, combination of m1A modification inhibitor thiram with PPARd antagonist had a synergetic inhibitory effect on tumour growth, and extended lifespan in PDC models from HCC patients with higher m1A levels, suggesting that targeting TRMT6/TRMT61A complex-m1A modification axis may be an effective method in liver cancer treatment.
This work was well done with great efforts by all collaborative groups. Without their support, we would not be able to reveal the critical role of m1A modifications in liver cancer development. In summary, our findings provide insights into the dynamic and regulatory role of m1A modifications in tRNA underlying hepato-oncogenesis. We hope our work will be of great interest to others and pave a way to develop more effective therapeutic strategies for HCC patients.
1 Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics, 2019. CA Cancer J Clin 69, 7-34 (2019).
2 Kaiser, J. The cancer stem cell gamble. Science 347, 226-229 (2015).
3 Wang, Y. et al. The long noncoding RNA lncTCF7 promotes self-renewal of human liver cancer stem cells through activation of Wnt signaling. Cell Stem Cell 16, 413-425 (2015).
4 Li, X. et al. Base-Resolution Mapping Reveals Distinct m(1)A Methylome in Nuclear- and Mitochondrial-Encoded Transcripts. Mol Cell 68, 993-1005 e1009 (2017).
5 Dominissini, D. et al. The dynamic N(1)-methyladenosine methylome in eukaryotic messenger RNA. Nature 530, 441-446 (2016).