A genome-scale CRISPR deletion screen in Chinese hamster ovary cells reveals essential regions of the coding and non-coding genome.

The biopharmaceutical sector relies on CHO cells to investigate biological processes and as the preferred host for production of biotherapeutics. Simultaneously, advancements in CHO cell genome assembly have provided insights for developing sophisticated genetic engineering strategies. While the majority of these efforts have focused on coding genes, with some interest in transcribed non-coding RNAs (e.g., microRNAs and lncRNAs), there remains a lack of genome-wide systematic studies that precisely examine the remaining 90 % of the genome and its impact on cellular phenotypes. This unannotated "dark matter" includes regulatory elements and other poorly characterized genomic features that may be potentially critical for cell behaviour. In this study, we deployed a genome-scale CRISPR screening platform with 112,272 paired guide RNAs targeting 14,034 genomic regions for complete deletion of 150 kb long sections. This platform enabled the execution of a negative screen that selectively identified dying cells to determine regions essential for cell survival. By using paired gRNAs, we overcame the intrinsic limitations of traditional frameshift strategies, which will likely have little or no effect on the non-coding genome. This study revealed 427 regions essential for CHO cell survival, many of which currently lack gene annotation or known functions. For these regions, we present their annotation status, transcriptional activity and annotated chromatin states. Selected regions, particularly those lacking all of the above, were individually deleted to confirm their essentiality. This work sheds a novel light on a substantial portion of the mammalian genome that has been traditionally difficult to investigate and therefore neglected. Notably, the fact that the deletion of some of these regions is lethal to cells suggests they encode critical regulatory functions. A better genome-wide understanding of these functions could open new avenues for engineering cells with improved bioprocess relevant properties.