Abstract
Long noncoding RNAs (lncRNAs) constitute a significant portion of the human transcriptome, yet their precise role in tumor cells remains elusive. In our study focusing on multiple myeloma (MM), we employed the CRISPR-Cas13d endonuclease's RNA-targeting capabilities to establish an innovative functional transcriptomics screening platform. This platform facilitated the comprehensive identification of tumor-promoting lncRNAs (tp-lncRNAs) in MM cells and provided mechanistic insights. Utilizing RNA-seq data from 319 MM patients, we identified 5,304 expressed lncRNAs. Subsequently, we designed and implemented a pooled lentiviral library incorporating nine single-guide RNAs (sgRNAs) per identified lncRNA, alongside 500 non-targeting sgRNAs as negative controls and nine sgRNAs targeting established protein-coding oncogenes in MM (e.g., IRF4) as positive controls. This library was introduced at a low multiplicity of infection (MOI < 0.3) across five MM cell lines, each expressing a Cas13d-GFP fusion protein. After three weeks, deep sequencing and MAGeCK assessed sgRNA depletion/enrichment in MM cells. Our findings revealed ~600 tp-lncRNAs upon which MM cell lines demonstrated significant growth dependency, constituting around 12% of the identified lncRNA transcriptome. Notably, this surpassed the previous estimation of 3–5%, with considerable overlap across MM cell lines. The dependency score of numerous tp-lncRNAs rivaled or exceeded that of the positive control IRF4. To validate dependencies, we employed a library of 3,500 sgRNAs to target 314 tp-lncRNAs in vitro and in xenograft models. This approach was also applied to a panel of six non-MM cancer cell lines, identifying a broad tumor-dependency on >59 lncRNAs and a specific MM-dependency on a smaller subset of lncRNAs. We explored the sub-cellular localization and site-of-function of tp-lncRNAs using a Cas13d variant accumulating either in the cytosol or in the nucleus, coupled with sub-cellular RNA-seq and single-molecule RNA FISH. Confirming the functionality and nuclear enrichment of most tp-lncRNAs, we identified >20 hits providing dependency in the cytosol. Our investigation also unveiled multiple transcript variants within the lncRNA transcriptome, originating from the same lncRNA genes but differing in sequences, structures, and likely functions. Crucially, our screening platform precisely identified tp-lncRNA isoform(s) conferring dependencies. Finally, we combined short-term Cas13d screens with single-cell RNA-seq in H929 MM cells. Analysis of over 20,000 cells for gRNA and transcript expression enabled us to delineate molecular pathways linked to individual perturbations of over 30 lncRNAs in a single experiment. The modulated pathways included therapeutically relevant ones, such as the unfolded protein response or mitochondrial homeostasis, further suggesting a potential therapeutic benefit from targeting tp-lncRNAs. In summary, our novel platform offers a comprehensive delineation of tumor dependency on lncRNAs in MM, providing valuable insights to guide further mechanistic inquiries.
Citation Format: Eugenio Morelli, Anil Aktas-Samur, Domenico Maisano, Claire Gao, Na Liu, Vanessa Favasuli, Marcello Turi, Pietro Folino, Liang Chen, Francesca Barello, Mariateresa Fulciniti, Annamaria Gulla, Kenneth Anderson, Mehmet Kemal Samur, Nikhil Munshi. Decoding Cancer Dependency and Molecular Pathways of Long Noncoding RNAs with a Novel CRISPR-Cas13d Platform [abstract]. In: Proceedings of the Blood Cancer Discovery Symposium; 2024 Mar 4-6; Boston, MA. Philadelphia (PA): AACR; Blood Cancer Discov 2024;5(2_Suppl):Abstract nr P37.