While it is now known that human chromosomes are pervasively transcribed, the relative contribution of the coding and lncRNA genomes to cell differentiation is unknown. Using genome-wide CRISPR interference (CRISPRi) screens in human pluripotent stem cells, we systematically assessed 18,905 protein-coding and 10,678 lncRNA targets for function in neural induction, identifying 419 coding and 201 lncRNA genes that regulate this step of development. Comparative multi-omic analyses discovered broad properties of coding and lncRNA genome function, such as an enrichment of lncRNA genes for roles in differentiation rather than proliferation. In contrast, coding genes were enriched for function in cell proliferation. CRISPRi single-cell RNA-Seq (Perturb-Seq) analysis resolved differentiation phenotypes at the level of specific cellular states, revealing additional distinctions in how coding and lncRNA genes regulate neural induction. These systematic studies demonstrated key functional differences between the coding and lncRNA genomes during this critical stage of neurodevelopment.

RNA plays a dual role as an informational molecule and a direct effector of biological tasks. The latter function is enabled by RNA's ability to adopt complex secondary and tertiary folds and thus has motivated extensive computational and experimental efforts for determining RNA structures. Existing approaches for evaluating RNA structure have been largely limited to in vitro systems, yet the thermodynamic forces which drive RNA folding in vitro may not be sufficient to predict stable RNA structures in vivo. Indeed, the presence of RNA binding proteins and ATP-dependent helicases can influence which structures are present inside cells. Here we present an approach for globally monitoring RNA structure in native conditions in vivo with single nucleotide precision. This method is based on in vivo modification with dimethyl sulfate (DMS), which reacts with unpaired adenine and cytosine residues9, followed by deep sequencing to monitor modifications. Our data from yeast and mammalian cells are in excellent agreement with known mRNA structures and with the high-resolution crystal structure of the Saccharomyces cerevisiae ribosome. Comparison between in vivo and in vitro data reveals that in rapidly dividing cells there are vastly fewer structured mRNA regions in vivo than in vitro. Even thermostable RNA structures are often denatured in cells, highlighting the importance of cellular processes in regulating RNA structure. Indeed, analysis of mRNA structure under ATP-depleted conditions in yeast reveals that energy-dependent processes strongly contribute to the predominantly unfolded state of mRNAs inside cells. Our studies broadly enable the functional analysis of physiological RNA structures and reveal that, in contrast to the Anfinsen view of protein folding, thermodynamics play an incomplete role in determining mRNA structure in vivo.

The flow of genetic information through transcription and translation—known as the central dogma of biology—enables all aspects of life, from the response of a solitary yeast cell to an environmental stimulus, to the intricate choreography guiding development of a single-celled zygote into a complex organism with functionally distinct tissues. Recent technological advances have provided tools to observe and/or perturb molecular processes underlying the central dogma with unprecedented resolution and precision. In this dissertation, I describe two cases where such tools were used to study and manipulate gene expression in eukaryotic cells.

First, I show that in budding yeast, under nutrient limiting conditions, the commonly used translation inhibitor cycloheximide induces rapid transcriptional upregulation of hundreds of genes involved in ribosome biogenesis. This generates a large pool of mRNAs that cannot be translated due to the presence of the inhibitor, leading to the appearance of strong translational regulation. This work provides a novel mechanistic interpretation for perplexing reports in the translation field, and hopefully serves to guide experimental design moving forward.

Second, I describe the development of allelic series of systematically compromised sgRNAs to titrate expression of human genes with CRISPR interference. Large-scale measurements of compromised sgRNA activities enable identification of empirically validated intermediate-activity sgRNAs and determination of the factors governing sgRNA activity using deep learning, facilitating construction of a compact sgRNA library to titrate expression of ~2,400 essential genes and a genome-wide in silico library. Staging cells along a continuum of essential gene expression levels using sgRNA series combined with rich single-cell RNA-seq readout reveals expression threshold-specific responses and gene-specific expression-to-phenotype relationships, thus highlighting such reagents as a general tool to titrate the expression of human genes, with potential applications ranging from tuning of biochemical pathways to identification of suppressors for diseases of dysregulated gene expression.

Coding sequences (CDSs) have traditionally been predicted on the basis of sequence and length; on any given mRNA the longest ATG-initiated CDS is annotated as the translated protein product. Ribosome profiling, a technique for sequencing regions of mRNA being decoded by the ribosome, enables a high-resolution view of translation and is thus a method for the data-dependent annotation of CDSs. When performed in the presence or absence of drugs that target the ribosome in stereotyped ways, ribosome profiling reveals the core features of translation: initiation, phased elongation, and termination.

We developed and applied a computational analysis pipeline that leverages all these features to identify CDSs systematically. Notably, our analysis is not biased by length, initiator codon, or overlapping CDS structure. We performed our experiments in bone marrow-derived dendritic cells undergoing lipopolysaccharide (LPS) stimulation. Cell type-specific transcript assemblies guided our analysis, and mass spectrometry corroborated our predictions. We identified translation of both annotated and unannotated CDSs.

Unannotated translation comprises previously unknown variants of annotated proteins and completely novel CDSs. We identified 829 N-terminal truncations and 609 N-terminal extensions, many of which affect localization signals or domain structure. We identified 5,215 CDSs translated from regions upstream of annotated CDSs; these upstream CDSs may regulate expression of their associated downstream CDS. We identified 472 new CDSs that arise from new transcripts or transcripts previously defined as non-coding. Finally, we found 989 translation events that occur within but out-of-frame relative to annotated CDSs.

We used ribosome density to quantify expression of all identified CDSs at various times after LPS stimulation (0, 0.5, 1, 2, 4, 6, 8, 9, 12 hrs). Many newly identified CDSs display significantly enhanced or repressed expression through the time-course, relative to mock-stimulated cells.

The genome sequencing efforts of the past decade have cataloged the universe of protein-coding and non-coding transcripts produced by the human genome. A central challenge now is to understand how these many thousands of genes act to mediate the vast array of cellular processes involved in either normal or disease states. To accomplish this, we require specific and scaleable tools to manipulate the expression of individual genes and measure their functional contribution in a given cellular context. In this thesis, I describe the initial development and subsequent refinement of genome-scale genetic screening technologies based on CRISPR-mediated interference and activation. I present the results of several screens for protein-coding genes that modulate cell proliferation and toxin susceptibility. Finally, I describe the application of CRISPR interference in the systematic repression of long non-coding RNA genes, a large class of genes of which very few genes have any known function. The screens described here identify nearly 500 lncRNA genes that modify cell growth in at least one of seven diverse cell lines, and furthermore highlight the exquisite cell type-specificity of lncRNA function.

Biological systems use proteins to drive just about every cellular process. Thus one would gain much insight into the physiology of an organism by knowing the catalog of proteins that are being actively expressed. Ribosome profiling (deep sequencing of ribosome-protected mRNA footprints) enables one to monitor protein production genome-wide. Here, I describe the application of ribosome profiling to the Gram-negative bacterium Escherichia coli, and show that translation control is pervasive. In addition, I review the regulatory mechanisms that govern transient ribosome pausing, which can influence a number of cotranslational processes, such as folding and localization of the nascent chain. I continue this discussion by showing that pausing is widespread, and mainly stems from transient interactions made between the anti-Shine-Dalgarno sequence of the 16S ribosomal RNA and the messenger RNA template, rather than the cellular concentrations of transfer RNA. Finally, I describe a new application of the ribosome profiling approach that enables one to monitor cotranslational processes in vivo. Here, I characterize the cotranslational action of the molecular chaperone trigger factor. This study unexpectedly revealed that trigger factor does not engage nascent chains until well after synthesis, contrary to in vitro findings, and that outer membrane porin proteins are the most prominent substrates of this chaperone.

Advances in DNA sequencing have provided the complete genome sequence for many organisms. However, understanding the functions of each gene product encoded by the genome remains a major challenge. This dissertation describes an improved strategy for identifying functional relationships between genes and the use of unbiased genetic and biochemical approaches to uncover the function of the ORM family of genes.

Large-scale measurement of genetic interactions has proven to be a powerful approach for functional genomics in the model organism Saccharomyces cerevisiae. The key to such analyses is comparing the effects of mutations to individual genes or pairs of genes on a phenotype of interest, such as growth rate. In the first part of my thesis, I developed a high-throughput, flow cytometry-based strategy for measuring growth rates in S. cerevisiae with improved accuracy. This assay made it possible to precisely quantify the contribution of each gene to growth and to measure genetic interactions with increased sensitivity. With high-precision data, I was able to resolve an enigma regarding the apparent prevalence of gene dispensability in yeast, to reveal new relationships between chromatin modifying factors, and to create a functional map of the proteasome.

In the second part of my dissertation, I used genetic interactions as a starting point to identify members of the conserved ORM gene family as negative regulators of sphingolipid synthesis. I found that Orm proteins form a complex with serine palmitoyltransferase, the first and rate-limiting enzyme in sphingolipid production, and that a phosphorylation-based regulatory pathway relieves the inhibitory activity of Orm proteins when sphingolipid production is disrupted. This feedback mechanism matches sphingolipid supply to cellular demand, with changes in ORM gene expression or mutations to their phosphorylation sites causing dysregulation of sphingolipid metabolism. Thus, given the essential roles of sphingolipids as structural components of membranes and as signaling molecules, Orm proteins serve a key role in cellular homeostasis. Finally, the functional conservation within this gene family and the recent identification of the human homolog ORMDL3 as a potential risk factor for childhood asthma raise the possibility that sphingolipid misregulation contributes to asthma development.

This work focuses on how the endoplasmic reticulum-associated degradation (ERAD) quality control machinery is able to distinguish terminally misfolded glycoproteins from folding intermediates and target these potentially toxic forms across the ER membrane for degradation. For luminal, glycosylated proteins, this discrimination depends both on the protein's folding status and on its glycosylation state, but much is still unknown about the mechanism.

My research centers on Yos9p, a recently identified protein that is essential for luminal ERAD in Saccharomyces cerevisiae. I demonstrated that Yos9p's predicted glycan-binding domain is necessary for ERAD and together with data showing that Yos9p interacts with ERAD substrates, these results implicated Yos9p as a critical sensor of the glycan-based degradation signal on misfolded substrates. Thus two important questions arose: how does Yos9p participate with the rest of the ERAD machinery to target substrates for degradation, and what is the glycan destruction signal that Yos9p recognizes? By affinity purifying Yos9p, I found that Yos9p, its binding partner, Hrd3p, and the chaperone Kar2p form the luminal components of a multi-protein complex that organizes key ERAD factors including the transmembrane ubiquitin ligase Hrd1p and cytosolic AAA ATPase Cdc48p. My characterization of this complex complemented the observation that Yos9p and Hrd3p each form a complex specifically with misfolded substrates independently of substrate glycosylation. My studies further suggested that Yos9p and Hrd3p act as gatekeepers that ensure that only legitimate substrates are degraded. Together, these findings indicate that Yos9p forms a luminal surveillance complex with Hrd3p to recruit misfolded proteins to the core ERAD complex and that Yos9p assists in a distinct sugar-dependent step necessary for degradation of correct substrates.

To characterize Yos9p's glycan specificity, I developed a protocol to make functional Yos9p from E. coli by refolding the protein in vitro. Frontal affinity chromatography determined that Yos9p recognizes N-linked glycans containing a terminal alpha 1,6-linked mannose. My in vivo studies then showed that this glycan signal is necessary for degradation. I also provide evidence that another ERAD factor, Htm1p, generates this specific glycan for recognition by Yos9p, adding another step to potentially increase the specificity of the recognition process.

The yeast prion [PSI+] is caused by the self-propagation of beta-sheet rich aggregates of the translation termination factor Sup35. Studies in this system have shown some of the most definitive proof of the protein-only nature of prions and have explained the strain phenomenon by showing that heritable information can be encoded by the conformation of a protein. The structural basis of how these various protein conformations lead to phenotypic differences has been difficult to discern primarily due to the technical difficulties of working with an aggregating protein. The use of non-perturbative stable isotopic labels and solid-state NMR (SSNMR) has been successful in characterizing some semi-ordered protein aggregates. However, the Gln/Asn-rich nature and low sequence complexity of Sup35's prion domain has prevented sequence specific information from being generated by SSNMR rendering this technique less useful. The work presented here was primarily aimed at generating a tool to allow for site-specific isotopic labeling of SupNM. I describe a general strategy for the semisynthesis of Gln/Asn-rich proteins utilizing optimized chemical synthesis, recombinant protein production, and native chemical ligation. I show that semisynthetic SupNM is capable of forming amyloid fibers that are competent to infect yeast. Utilizing this method will allow for precise placement of isotopic labels and enable residue-specific chemical shift assignments in SSNMR with the goal of generating high-resolution molecular models of the [PSI+] prion.