The function of most of the noncoding genome has not been experimentally tested.
The human genome produces tens of thousands of long noncoding RNAs (lncRNAs), transcripts more than 200 nucleotides in length that do not code of proteins.
There are more than twice as many lncRNA genes as coding genes. It is currently not possible to predict lncRNA function from the genomic sequence.
Which lncRNAs have important biological function?
Only ~2% of the genome codes for protein (left), but ~80% of the genome is transcribed into RNA (right).
CRISPRi for high-throughput screens in human cells
Single guide RNA (sgRNA) can be synthesized to knockdown expression of any gene, coding or noncoding. Large sgRNA libraries can interrogate the function of tens of thousands of genes in a single experiment.
We developed CRISPR interference (CRISPRi) as a tool for systematic, genome-scale screens of lncRNA gene phenotypes (Liu et al., Science 2017). By targeting knockdown of over 16,000 lncRNA genes in seven different cell lines – six cancer cell types and human induced pluripotent stem cells – we identified 499 lncRNA genes required for cellular growth.
Remarkably, 89% of these functional lncRNAs are required for the growth of only one cell type, with lncRNA knockdown perturbing complex transcriptional networks in a highly cell type-specific manner.
Thus, in addition to greatly increasing the number of functional lncRNA loci, these results highlight the exquisite cell type-specificity of essential lncRNAs, which has important implications for both basic biology and disease.
Finding lncRNA therapeutic targets
Targeting lncRNAs might kill cancer cells while sparing the normal cells around the tumor. Glioblastoma (GBM) is fatal in nearly all patients. Fractionated radiotherapy is a part of the standard-of-care for GBM.
Ideally, a new therapeutic would sensitize GBM cells to radiation while sparing normal brain cells.
We have used CRISPRi to discover lncRNA targets that sensitize GBM cells to clinically relevant doses of radiation (Liu, Malatesta, et al., Genome Biol 2020).
Testing in "mature" brain organoids (MBOs) confirms selective killing of tumor cells.
We continue to develop the most promising target, called lncGRS-1.
Some ongoing research...
- Dual coding and lncRNA screens for human neural induction and neurogenesis.
- Identifying functional subclasses of lncRNAs
- Machine learning to identify predictors of lncRNA function.
