Identification and functional characterization of long non-coding RNAs in development and disease
- Author(s): Liu, Siyuan John
- Advisor(s): Lim, Daniel A
- et al.
The human genome produces tens of thousands of long non-coding RNA (lncRNAs), transcripts larger than 200 nt that do not make proteins. Select lncRNAs are established regulators of gene expression in developmental processes and in the etiology of diseases. However, the functions of the vast majority of lncRNAs are unknown. Here we utilize high throughput descriptive and functional genomics to characterize and understand the roles of lncRNAs in development of the human brain and in cancer biology. Starting with deep RNA sequencing (RNA-seq) of developing brain tissues, we comprehensively annotated lncRNA transcript structures in the cerebral cortex, revealing thousands of novel lncRNAs, many of which were antisense or non-polyadenylated. We then applied single cell RNA-seq to quantitatively measure the transcriptomes of individual cells, revealing cell type-specific expression of lncRNAs and surprisingly high abundance of certain lncRNAs (e.g. DLX6-AS1) that appear to be low abundance at the population level, due to cellular heterogeneity in the brain. Combination of single cell RNA-seq and exome-seq were then applied to not only measure RNA expression heterogeneity, but also to map the mutational phylogenies of glioblastoma tumors, revealing a novel deletion in PDGFRA that confers survival advantage. We then developed a CRISPR interference (CRISPRi) platform to systematically repress 16,401 lncRNA loci across seven cancer and non-cancer cell lines, identifying 499 lncRNAs necessary for robust growth and proliferation. The functions of these lncRNAs were exquisitely cell type-specific and correlated with complex transcriptional networks and chromatin architecture. Finally, we applied CRISPRi screening of lncRNAs in combination with ionizing radiation therapy and identified over 200 lncRNAs whose knockdown sensitizes glioblastoma cells to radiation induced cell death. As much as these efforts advance our knowledge of lncRNA biology in development and in disease, they also raise exciting new questions for future inquiry.