Cells must respond to a wide range of external stimuli by altering their gene expression programs. The contribution to these responses of changes to transcription initiation has been studied extensively, but subsequent mRNA processing steps can also be regulated. In this work, we have investigated the potential for regulation at the level of pre-mRNA splicing. We found that the pre-mRNAs for a majority of the ribosomal protein genes (RPGs) rapidly accumulated in response to amino acid starvation. By contrast, most RPG pre-mRNAs were rapidly degraded by the nuclear exosome in response to a variety of stresses that inactivated TORC1. These responses required casein kinase 2. In the case of amino acid starvation, the native promoter of an RPG, but not its native intron, was also required for the response. This suggested that the mechanism of splicing inhibition following amino acid starvation involved modulation of the co-transcriptional dynamics of spliceosome recruitment. Consistent with this model, we found that the inhibition of pre-mRNA splicing was accompanied by a rapid accumulation of both U1 snRNP and RNA polymerase II at the 3' ends of many ribosomal protein genes, suggesting that U1 release was inhibited. This inhibition of U1 release was insufficient to fully commit the RPG pre-mRNAs to eventual splicing however, as single molecule RNA FISH revealed that the accumulating pre-mRNAs localized to the cytoplasm during amino acid starvation. These results suggested a high degree of interaction among transcription, splicing, 3' end processing, and mRNA export, and we carried out a high throughput genetic screen to identify interactions among factors and complexes involved in mRNA processing. This screen revealed previously unknown connections between a proteasome component (Sem1) and mRNA export, and between a component of the COP9 signalosome and pre-mRNA splicing. We hope this dataset will also continue to be useful as we and others continue efforts to elucidate the interconnected regulatory mechanisms that contribute to control of gene expression at the level of mRNA processing.

The spliceosome is a dynamic ribonucleoprotein machine that identifies and removes introns from pre-mRNA molecules. It is composed of five small RNAs and over 100 proteins that assemble de-novo on each intron, through an ordered cascade of recognition, rearrangement, and re-recognition events. At its catalytic core, the spliceosome is thought to be a ribozyme, like the ribosome and group II introns. What then are the functions of the many protein components of the spliceosome? One group of proteins, the family of DExD/H-box ATPase, catalyzes conformational rearrangements to enhance fidelity and drive the ordered cascade of splicing events. Regulation of these events is critical for proper spliceosome function. One particularly interesting protein, Prp8, is thought to affect the structure of the catalytic core of the spliceosome and also serve as a master regulator of spliceosome assembly and activity. Yet how Prp8 interacts with other components of the spliceosome, to regulate their activity or to affect the structure of the catalytic core, is poorly understood.

We began by exploring how Prp8 regulates Prp28, the DEAD-box ATPase that promotes U1 release from and U6 association with the 5' splice site during spliceosome assembly. We found that a cluster of mutations in the N-terminus of Prp8 suppressed the growth defect of the cold-sensitive prp28-1 allele. Surprisingly, Prp28 was required not only for U1 release but also for earlier steps including ATP-independent formation of commitment complex 2. The prp28-1 suppressors in Prp8 also suppressed defects in early this ATP-independent step, suggesting that Prp8 and the U5 snRNP also participate in the very earliest steps of spliceosome assembly (Chapter 1). Next we explored the dynamics of commitment complex 2 formation and its dependence on Prp28 using single molecule FRET to probe pre-mRNA conformations (Chapter 2).

Prp8 is critical for splicing catalysis as well as assembly, and it sits at the catalytic core of the spliceosome. Two classes of Prp8 alleles have been characterized that bias the spliceosome toward two opposing conformations. We propose a model, based on comparisons to the group II ribozyme, that these conformations represent catalytic and transitional (rearrangement) conformations of the spliceosome. We used a cluster of mutations that stabilize two different structures in Prp8 to investigate how a structural toggle in Prp8 affects the catalytic and transitional conformations of the spliceosome. Supporting our model, we found that Prp8 alleles predicted to stabilize the catalytic conformation were error prone and catalyzed in the 1st step of splicing efficiently in vitro, while alleles predicted to stabilize the transitional conformation were hyper-accurate and inefficient at the 1st step of splicing (Chapter 3).

Both spliceosome assembly and catalysis are points at which the fidelity or activity of the spliceosome could be regulated. My goal in this research has been to understand how individual steps are controlled by proteins such as Prp28 and Prp8 in hopes that this will give us a better context in which to study how cells could regulate these components to regulate splicing. New structural studies are giving the field the power to generate more informed hypotheses, based on structure, as a starting point for future genetic and biochemical studies.

Splicing of eukaryotic pre-mRNAs often occurs co-transcriptionally; however, the components of the co-transcriptional environment that influence splicing remain incompletely defined. The work presented here discusses the effects of altering histone modifications, in particular H2B ubiquitination, and RNA Polymerase II (RNAPII) elongation rate on splicing in Saccharomyces cerevisiae.

To begin our exploration into the effect of the co-transcriptional environment on splicing, we interrogated the genetic, functional, and biochemical interactions of Npl3, an SR-like protein in S. cerevisiae that promotes co-transcriptional spliceosome assembly. Among many chromatin and transcription factors, we found that Npl3 genetically interacts with the histone H2B ubiquitination machinery and went on to show that strains lacking H2B ubiquitination have a mild splicing defect (Chapter 1). We subsequently capitalized on the observation that H2B ubiquitination-dependent processes are sensitive to high temperatures, and observed splicing defects upon inhibition of H2B ubiquitination or related modifications, H3K4 and H3K36 methylation, with each modification exerting gene-specific effects. Furthermore, semi-quantitative mass spectrometry on purified nuclear mRNPs and chromatin immunoprecipitation analyses on intron-containing genes indicated that H2B ubiquitination stimulates recruitment of the early splicing factors onto nascent RNAs (Chapter 2).

It is challenging to nail down the precise mechanisms by which splicing responds to histone modifications - direct physical interactions between the modification and the spliceosome would offer locus- and modification-specific recruitment to promote splicing efficiency, but an additional variable, RNAPII elongation rate, had been shown to be a key player in metazoans. We therefore interrogated the effect on splicing of altering RNAPII elongation rate in S.cerevisiae, using an allelic series of point mutants that cause fast and slow elongation in vitro. Our splicing microarray experiments revealed that RNAPII speed and splicing efficiency are anti-correlated: at many genes, increased elongation rate caused decreased splicing efficiency, while decreased elongation rate increased splicing efficiency (Chapters 3 and 4).

Taken together, our data support the growing notion that histone modifications and RNAPII elongation rate exert gene-specific effects and potentially offer multiple locus-specific mechanisms for modulating splicing in a dynamic context.

During eukaryotic gene expression, non-coding intron sequences are removed from the protein-coding exon sequences of pre-messenger RNA (pre-mRNA) by the process of splicing. Splicing is carried out by the spliceosome, a macromolecular ribozyme that assembles de novo on every intron. It has been established that the spliceosome assembles co-transcriptionally while the nascent pre-mRNA is attached to the elongating RNA polymerase II (RNAP II). However, mechanisms that enforce the coupling of splicing to transcription and the chromatin landscape remain poorly understood.

Here, I describe our work to gain further insight into the regulation of co-transcriptional pre-mRNA splicing by employing genetic and molecular biology approaches. Using the traditional splicing model organism Saccharomyces cerevisiae, we interrogated the mechanism of a rapid splicing block at the ribosomal protein genes (RPGs) that occurs in response to amino acid starvation stress. By creating chimeric gene constructs, we show that an RPG promoter sequence element, and not the intron sequence, is both necessary and sufficient to confer a transcript-specific splicing response. This study reveals a mechanism whereby in response in stress RNAP II becomes de-enriched near the 3´splice site and early splicing factor recruitment is delayed, consistent with an environmentally-responsive splicing-dependent polymerase checkpoint, and underscoring the highly coordinated nature of splicing and transcription.

We further expanded our studies of splicing regulation into Schizosaccharomyces pombe, a emerging model organism that shares many splicing features with mammalian systems, such as abundant non-consensus splice sites, multi-intron containing genes, and conserved trans-acting splicing factors. In our work with the S. pombe system, we developed techniques for probing splicing factor association with chromatin. Starting from a genome-wide genetic screen, we characterized interactions between the spliceosome and Swr1 complex, which deposits the highly conserved histone variant H2A.Z. We identified H2A.Z’s requirement in efficient splicing at introns with weak splice sites, supporting a growing hypothesis of chromatin-dependent splicing checkpoints. Taken together, these studies reveal an unappreciated role of the promoter and its proximal nucleosome composition in transcript-specific pre-mRNA splicing regulation. We expect these results to provide foundational insights into the interconnectedness of gene expression processes and have important implications in human disease.

Eukaryotic genes are littered with intervening sequences, or introns, that are transcribed, but must be precisely excised from a messenger RNA before it can be properly translated into protein. While introns were once regarded as "junk DNA," they are not inconsequential. However, we currently lack the knowledge to accurately predict the functions, if any, of individual introns in any organism. Here, I describe efforts to better understand the evolution and function of introns, with the vision that we will soon be able to identify important introns and predict their functions.

In Chapter 2, my colleagues and I describe an unexpected consequence of intron presence on chromatin structure, which suggests that introns have a broader influence on the biology of eukaryotes than previously appreciated. By analyzing published surveys of nucleosomes and 41 chromatin marks in humans, we show that 5' intronic and 3' exonic regions of active genes are differentially marked by characteristic chromatin marks, thus contributing substantially to the patterns of histone modifications within active genes. Intriguingly, these modification patterns were stable despite dramatic changes in the frequency of splice site usage at two alternative spliced genes. Thus, similar to promoter marks, which are relatively stable to differences in productive transcription, we propose that intronic and exonic chromatin marks reflect exon definition, rather than splicing per se.

In Chapter 3, I describe my work to better understand why certain introns persist in eukaryotic genomes. Using comparative genomics, I show that the ribosomal protein genes of Saccharomyces cerevisiae have greatly resisted intron loss. Mimicking the effect of intron loss with directed mutagenesis, I perform experimental tests that demonstrate that these introns do not promote gene expression, but rather, provide a means for regulation. Specifically, I show that the genes encoding ribosomal protein S9, both in S. cerevisiae and Drosophilia melanogaster, autoregulate in an intron-dependent manner. Lastly, I summarize published gene expression data from diverse animals, which suggest that multiple forms of alternative splicing have evolved to autoregulate S9 gene expression. Thus, I propose that the introns of eukaryotic genes persist, in part, due to their propensity to evolve regulatory function.

The spliceosome is one of the cells largest and important molecular machines and yet it remains as one the least understood systems. This is a result of many experimentally challenging features inherent to the biology of splicing. In addition to having the largest number of unique proteins and RNAs associated with one process, the spliceosome has a complex assembly process that is both conformationally and compositionally dynamic. Classical genetics, much of it done in the Guthrie lab, revealed which proteins and RNAs were required for the process of splicing, but the molecular details of their interactions is still relatively unknown. In the past 10 years new biochemical tools have been developed, such as smFRET and metal substitution studies, which have increased our abilities to probe the structure and activity of the spliceosome. These techniques and others outlined in this proposal will be the foundation for the next ten years of research on the spliceosome. I believe that through these efforts and building on recent breakthroughs, the molecular mechanism of pre-mRNA splicing will be determined. These are truly exciting times to be studying the spliceosome.

Eukaryotic mRNAs must undergo a diverse set of complex processing events before the messages that they carry can be competent for translation. Each of these stages in the life of an mRNA represent potential opportunities for regulating the fate of a message. In this work we have explored the relationship between changing cellular needs, induced by alterations in environmental conditions, and the process of splicing. Across a wide range of different environmental contexts we see evidence of rapid, transcript-specific changes in splicing efficiency. The splicing of transcripts can be both up-regulated and down-regulated in response to environmental cues. In some cases large numbers of transcripts are affected, in others only defined subsets. In total, our data reveal that splicing is a remarkably dynamic and flexible mechanism for post-transcriptional gene regulation in yeast.