Ribosome rescue factor PELOTA modulates translation start site choice and protein isoform levels of transcription factor C/EBPα [CRISPRi]

Translation initiation at alternative start sites can dynamically control the synthesis of two or more functionally distinct protein isoforms from a single mRNA. Alternate isoforms of the hematopoietic transcription factor C/EBPα produced from different start sites exert opposing effects during myeloid cell development. This alternative initiation depends on sequence features of the CEBPA transcript, including a regulatory upstream open reading frame (uORF), but the molecular basis is not fully understood. Here we identify trans-acting factors that affect C/EBPα isoform choice using a sensitive and quantitative two-color fluorescence reporter coupled with CRISPRi screening. Our screen uncovered a role for the ribosome rescue factor PELOTA (PELO) in promoting expression of the longer C/EBPα isoform, by directly removing inhibitory unrecycled ribosomes and through indirect effects mediated by the mechanistic target of rapamycin (mTOR) kinase. Our work provides further mechanistic insights into coupling between ribosome recycling and translation reinitiation in regulation of a key transcription factor, with implications for normal hematopoiesis and leukemiagenesis. Sublibrary CRISPR-dCas9 knockdown (CRISPRi) screens were performed in two color reporter cell lines to identify trans-factors that drive translation start site choice on CEBPA. CRISPRi sublibrary screens were performed using four, compact BFP-tagged CRISPRi sublibraries containing 5 sgRNAs per TSS (Addgene, #83971-3 and #83975) expressed in the pCRISPRi-v2 expression vector (Addgene, #84832). Plasmid sublibraries were separately packaged in HEK 293T Lenti-X cells and transduced into the CRISPRi dual color reporter line at an MOI < 1 where the percentage of transduced cells by BFP expression after 2 days post-transduction was 20%-30%. At 2 days post-transduction, we performed fluorescence activated cell sorting (FACS) using an Aria Fusion (BD Biosciences) to select for cells expressing BFP. Cells with the highest (20%) BFP expression were collected and recovered in RPMI 1640 for 6 days post-FACS. Approximately 10 million cells were collected per sublibrary, maintaining an average sgRNA coverage of at least 500 cells per sgRNA. At 6 days post-BFP selection, cells were again sorted using a FACS Aria Fusion based on the ratio of green/red fluorescence from our CEBPA dual color reporter line. Approximately 40 million cells per sublibrary transduction were sorted into four, distinct green/red bins (20-25% of cells in each bin), with each bin containing 8-10 million cells to ensure an average sgRNA/cell coverage of at least 500. Genomic DNA was immediately harvested from these cells using the DNeasy Blood and Tissue kit (Qiagen, 69504) and sgRNA fragments were isolated by SbfI (New England Biolabs) restriction digestion and Ampure bead size selection then amplified by PCR for deep sequencing using custom sequencing primers. The sgRNAs were sequenced on an Illumina HiSeq-4000 using custom primers.