Systematically profiling the effects of genetic perturbations is a powerful approach that has revealed the molecular basis for a wide range of biological phenomena. The simple, programmable DNA recognition of CRISPR-Cas9 enables genome-wide genetic analysis in human cells and many other systems. Cas9 is guided by a short RNA to a complementary sequence in the genome, where it can introduce mutations or alter gene expression. Pooled libraries of guide RNAs (gRNAs) that individually target each gene in the genome allow us to introduce genetic perturbations systematically into a population of cells. A key challenge is measuring the phenotypic effects caused by individual guides in these pooled libraries and linking these phenotypes back to the associated gRNA, thereby finding the gene that is responsible. Gene expression changes provide a clear and sensitive phenotypic measure for many cellular processes. We sought a general approach to profile how the expression of a particular gene of interest changed when other genes were perturbed. We began with a library of gRNAs, each disrupting one gene, and linked these guides with an expression reporter containing a guide-specific nucleotide barcode. gRNAs that alter reporter expression will change the abundance of the expressed RNA barcode specifically associated with that guide. Deep sequencing of these expressed barcodes quantifies each of these guide-specific reporter expression effects individually within a pooled, complex population. We have implemented this strategy by combining CRISPR interference (CRISPRi) with barcoded expression reporter sequencing (CiBER-seq). CiBER-seq profiling in yeast recapitulated known regulation across a range of promoters. Notably, interrogation of the integrated stress response (ISR) revealed both well-studied and uncharacterized modes of regulation. While uncharged tRNAs activated the ISR, tRNA insufficiency also activated the reporter, independent of the uncharged tRNA sensor. By uncovering alternate triggers for ISR activation, we illustrate how precise, comprehensive CiBER-seq profiling provides a powerful and broadly applicable tool for dissecting genetic networks. We extend the CiBER-seq paradigm to interrogate mammalian systems and to explore new protein-centric biological questions.

Translation initiation at alternative start sites has the potential to yield functionally and structurally distinct protein isoforms that can play decisive roles during various cellular processes including cell fate commitment. It further serves as a dynamic point of control in gene expression, enabling a rapid response to changes in the cellular environment through the translation of regulatory elements called upstream open reading frames (uORFs). Re-initiation after the translation of uORFs has been shown to regulate start site selection for downstream protein-coding genes, yet the molecular factors and mechanism underlying this choice has been largely uncharacterized. Here we identify the trans-acting factors that regulate start site usage in the key developmental transcription factor CEBPA using a sensitive and quantitative two-color fluorescence reporter coupled with CRISPRi screening. Our screen uncovered factors that have not been previously linked to re-initiation as well as recovered known regulators of re-initiation and CEBPA expression. In particular, we uncover a role for the ribosome recycling factor, Pelota (PELO), in the control of CEBPA translation. Using ribosome profiling and fluorescent based reporters, we provide evidence suggesting that PELO promotes long isoform expression by disfavoring CEBPA uORF translation in a manner that is partially dependent on mTOR activation. Our work provides further mechanistic insights into how ribosome recycling and translation re-initiation are coupled to regulate a major transcription factor with important implications for cell fate control.

At all stages of a messenger RNA’s lifecycle, it is covered in RNA-binding proteins. These proteins regulate an RNA transcript’s splicing and processing in the nucleus, its export from the nucleus into the cytoplasm, its localization and translation in the cytoplasm, and its eventual turnover and decay. Despite knowing the identities of roughly 700 RNA-binding proteins in budding yeast, the role in RNA regulation that many of these proteins perform remains unclear. Here we present two studies that are aimed at the functional characterization of proteins that regulate post-transcriptional gene expression. In the first study, we devised a high-throughput tethering assay for the characterization of proteins on a proteome-wide scale. This novel assay provides domain-level resolution for the functional regions of proteins and identifies their regulatory activity in a quantitative manner. In the second study, we characterized the yeast RNA-binding protein Mrn1p and found that it is a dynamic regulator of post-transcriptional regulation that functions through mRNA turnover. Mrn1p is especially important in linking cell wall biogenesis with mitochondrial homeostasis, and it regulates these two cellular compartments in a manner that is responsive to carbon source and cell stress. Together, we present two studies that provide new functional information about yeast RNA binding proteins, with broad implications for a better understanding of post-transcriptional gene expression.

RNA localization is a fundamentally important regulatory mechanism for the control of gene expression. I developed an ​in vivo ​subcellular RNA proximity labeling technique called APEX-Seq to monitor RNA localization and organization in the cell. In my approach, a proximity labeling enzyme is fused to a query protein and rapidly (<1 minute) biotinylates nearby RNAs ​in vivo upon addition of a labeling substrate, allowing their subsequent purification and analysis. We show first that APEX-Seq can distinguish the localization of transcripts to different regions of the cell. Further, we can detect the enrichment of specific transcripts in proximity to translation initiation proteins, showing that our system can capture more subtle, RNA-protein localization patterns. Finally, we present a high-resolution time course of protein and RNA condensation into stress induced RNA granules. A powerful aspect of APEX-Seq is the ability to match the spatial transcriptome with quantitative spatial proteomics, allowing for a more complete picture of the spatial landscape of the cell. Additionally, I rationally engineered a split APEX enzyme for conditional spatial proteomics and transcriptomics. I found evidence of unique translation initiation complexes through the use of split APEX on the initiation factors eIF1A/eIF4H, and eIF4A/eIF4B. Overall, these tools open the door for comprehensive and high throughout spatial transcriptomics, and proteomics with subcellular resolution.

Regulated changes in translation initiation control global protein synthesis to maintain cellular homeostasis, especially in times of nutrient deprivation and stress. The cap-binding protein eIF4E serves an integral role in translation initiation as the earliest contact between mRNAs and translation initiation machinery. The extent to which the cap-binding complex contributes to differential translation, and the mechanism underlying any specificity, remains poorly understood. Throughout this body of work, we aim to better understand the contribution of eIF4E to governing translational choice, maintaining cellular homeostasis and promoting stress responsive gene expression programs in S. cerevisiae.

In chapter 2, we investigated the effects of acute depletion of the essential cap-binding protein eIF4E on global post-transcriptional gene regulation. Surprisingly, cells nearly completely depleted of eIF4E exhibited only modest effects in cell growth and protein synthesis. Likewise, there were minimal changes in the selectivity of translation in response to eIF4E depletion. The strongest gene-specific effects of eIF4E depletion were induction of genes involved in the catabolism of aromatic amino acids. Further investigation revealed that the upregulation of these genes arose as secondary effects of reducing protein biosynthesis and the resultant rebalancing of amino acid pools. Furthermore, reductions in eIF4E activity caused translational activation of GCN4, a key regulator of stress responsive translation that is typically induced by amino acid starvation. Interestingly, translation of GCN4 occurred through a non-canonical mechanism instead of relying on the well-established effects of eIF2α phosphorylation. Additionally, we saw translational repression of PCL5, a negative regulator of Gcn4, in eIF4E depleted cells, which may have further contributed to imbalanced amino acid pools. This work provided new insights into the regulation of Gcn4 activity through Pcl5-mediated feedback control.

In chapter 3, we dissected the mechanisms by which cells robustly adapt to translation initiation stress caused by depletion of the cap-binding protein eIF4E. We assayed this adaptation by performing a synthetic genetic interaction screen. The screen uncovered a critical role for the non-canonical initiation factor eIF2A, encoded by YGR054W in yeast. Surprisingly, eIF2A acted to suppress aberrant translation during eIF4E depletion, despite previous evidence suggesting its role in initiating translation. Taken together, these findings underscore the complexity of the coordination between protein synthesis and amino acid homeostasis. This work enhances our understanding of the adaptive cellular response to acute depletion of a critical translation initiation factor. Additionally, the dysregulation of eIF4E and metabolism are hallmarks of cancer, emphasizing the importance of understanding the underlying mechanisms governing its action to develop effective therapeutic strategies.