UC Berkeley

A majority of metazoan genes contain introns in their primary transcripts (pre-mRNA) that require removal by the spliceosome--a cellular complex composed of protein and RNA. Upon removal of introns from the primary transcript, the remaining exonic portion of the transcript is spliced together. It is essential to remove the correct intronic portion of a primary transcript in order to produce the desired product, typically a protein-coding mRNA. Pre-mRNAs are alternatively spliced when different intron boundaries are used by the spliceosome, thus creating different mRNA products. Alternative splicing allows for an additional step of gene regulation by producing transcript isoforms that can be differentially processed in a particular tissue or developmental time point. Alternative splicing is primarily regulated by RNA binding proteins that bind to pre-mRNA and act to recruit or inhibit the spliceosome at specific splice sites. A central aim of this work is to gain a better understanding of splicing regulation by the identification and characterization of protein regulators of splicing and cis-acting splicing regulatory sequences in the model organism, Drosophila melanogaster.

To identify splicing regulatory elements, many previous studies in vertebrate genomes have used computational methods. In collaboration with Anna I. Podgornaia, I applied such an approach to predict splicing regulatory elements in Drosophila melanogaster and compared them with elements found in vertebrates. I identified 330 putative splicing enhancer sequences enriched near weak 5' and 3' splice sites of constitutively spliced introns. I found that a significant proportion (58%) of D. melanogaster enhancers were previously reported as splicing enhancers in vertebrates. To provide additional evidence for the function of the intronic splicing enhancers (ISEs), I identified intronic hexamers significantly enriched within sequences phylogenetically conserved among 15 insect species. This analysis uncovered 73 putative ISEs that are also enriched in conserved regions of the D. melanogaster genome. The functions of nine enhancer sequences were verified in a heterologous splicing reporter by Julie L. Aspden, demonstrating that these sequences are sufficient to enhance splicing in vivo. Taken together, these data identify a set of predicted positive-acting splicing regulatory motifs in the Drosophila genome and highlight those regulatory sequences that are present in distant metazoan genomes.

To identify and characterize splicing regulators, collaborators and I have combined RNAi and RNA-Seq to identify exons that are regulated by 58 known or putative splicing regulators. To identify and quantify alternative splicing events from RNA-Seq data, I developed the JuncBASE (Junction Based Analysis of Splicing Events) software package. For a pilot study, I identified 404 splicing events significantly affected upon depletion of pasilla. Preliminary analysis showed 879 splicing events affected by at least one of the 57 other proteins. The sequence regions upstream and within Pasilla-repressed exons and downstream of Pasilla-activated exons are enriched for YCAY repeats, which is consistent with the location of these motifs near regulated exons of the mammalian ortholog, Nova. Thus, the RNA regulatory map of Pasilla and Nova is highly conserved between insects and mammals despite the fact that the pre-mRNAs that are regulated by Pasilla and Nova are almost entirely non-overlapping. This observation strongly suggests that the regulatory codes of individual RNA binding proteins are nearly immutable, yet the regulatory modules controlled by these proteins are highly evolvable. I also present RNA regulatory maps for the four hnRNP proteins: hrp36, hrp38, hrp40, and hrp48.

Lastly, I examine splicing regulation throughout the life cycle of D. melanogaster. Using transcriptome data from 30 developmental time points produced by collaborators from the modENCODE Consortium, I identified a total of 23,859 alternative splicing events in Drosophila, taking into account all transcript information from D. melanogaster annotations, short sequenced reads (Illumina RNA-Seq), sequenced cDNA, long RNA-Seq reads (454 RNA-Seq) from adult flies, and short read sequences of rRNA-depleted RNA from embryonic time points. I observed that 60.7% of intron-containing genes in D. melanogaster are alternatively spliced. Using only the Illumina RNA-Seq reads throughout development, 21,216 splicing events were expressed and 13,951 events were differentially spliced in at least one time point. I also observed exons with similar patterns of splicing changes throughout development as well as sex-biased alternative splicing. Integrating information from our pasilla study, I also observed correlations of pasilla gene expression with alternative splicing changes of its target exons throughout development.