Poster abstracts
Poster number 139 submitted by Rajneesh Srivastava
Functional annotation of the human protein-RNA interaction network via transcriptome-wide pooled CRISPR/Cas9 screens to edit the RBP binding sites
Rajneesh Srivastava (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI), Aarthi Ramakrishnan (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI), Seyedsasan Hashemikhabir (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI), Sasank Vemuri (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI), Quoseena Mir (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI), Sarath Chandra Janga (Department of BioHealth Informatics, School of Informatics and Computing, IUPUI)
Abstract:
RNA-binding proteins (RBPs) bind to RNAs at specific recognition sites and decide the post-transcriptional fate of the targets within the cell. Although existing resources such as ENCODE, have documented millions of binding sites (BS) of 123 RBPs across cell types, the biological significance of these BS, is still lacking. Recently, we built SliceIt (https://sliceit.soic.iupui.edu/), an in silico library of ~4.8 million unique sgRNA (or single guide RNA) targeting all possible BS of 123 RBPs from eCLIP experiments in HepG2 and K562 cell lines from ENCODE project. It facilitates the easy navigation of RBP binding sites along with sgRNAs, SNP and GWAS annotations enabling researchers to conduct CRISPR experiments for studying the phenotypes associated with RBP binding sites. We designed a massive CRISPR/ Cas9 pooled screen for transcriptome-wide perturbation of RBP binding locations in HepG2 and K562 cells. We selected 23013 sgRNAs targeting 14256 unique binding (~52% exonic and ~48% intronic) sites of 108 RBPs in 4111 genes (including 254 lncRNA genes) with 89% of the BS targeted by atleast 2-8 sgRNAs. We also included ~134 control non-targeting sgRNA library in the screen. Pooled screening experiments were accomplished in HepG2 and K562 cell lines in replicates, resulting in the generation of ~50 million sgRNA reads (with ~99.8 correct sgRNAs representation) per replicate per cell line using Illumina Hi-seq followed by deep RNA-sequencing to generate 200 million reads per cell line of control and edited cells respectively. In a preliminary analysis of this dataset, we observed 68,323 and 87,319 exons were found exhibiting ±1.5 fold change in expression between control and edited HepG2 and K562 cells respectively. We anticipate this resource to be the first genome-wide CRISPR screen that will facilitate understanding and mapping the functional protein RNA interaction networks across human cell types.
References:
1. Van Nostrand, E. L. et al. Nature methods, doi:10.1038/nmeth.3810 (2016).
2. Dixit, A. et al. Cell. doi:10.1016/j.cell.2016.11.038 (2016).
3. Vemuri S et al. Methods. doi: 10.1016/j.ymeth.2019.09.004 (2019).
4. Datlinger, P. et al. Nature methods. doi:10.1038/nmeth.4177 (2017).
5. Rauscher, B. et al. Nucleic Acids Research. doi:10.1093/nar/gkw997 (2017).
6. Davis, C. A. et al. Nucleic Acids Research. doi:10.1093/nar/gkx1081 (2018).
Keywords: RNA binding proteins, in silico guideRNA library, pooled CRISPR screening