Abstract
Abstract 978
Alternative pre-mRNA splicing plays a major role in development and differentiation by re-modeling the transcriptome to generate mRNAs that encode the biologically appropriate cell type-specific proteome. Earlier studies employing RT-PCR and exon microarrays demonstrated a small number of splicing changes during erythroid differentiation, one of which (in protein 4.1R) is critical for mechanical stability of the membrane skeleton. Here we report that the landscape of splicing changes executed during terminal erythropoiesis is far more extensive and highly dynamic, ultimately affecting the expression of many more proteins than previously recognized. Highly purified populations of FACS-sorted cells representing erythroblasts at distinct stages during terminal erythroid differentiation from proerythroblasts to orthochromatic erythroblasts of both human and mouse origin were used as the source of RNA for RNA-seq analysis. In total, hundreds of millions of sequence reads were obtained from three biological replicates for four (mouse) or five (human) cell populations, and reads were aligned to the Ensembl-annotated transcriptome using the Bowtie aligner. Transcript-level estimates were obtained using the streaming transcript abundance estimation tool, eXpress, expression of individual exons in “exon-inclusion” isoforms relative to total isoforms was represented as Ψ (psi), or percent spliced in, and statistical significance estimates adjusted for multiple comparisons by the Benjamini-Hochberg method. Thousands of alternative splicing events were predicted in genes with diverse functions in transcription, RNA processing, protein synthesis, membrane receptors, cytoskeletal structure, etc. Initial RT-PCR studies indicate that a high proportion of predicted alternative splicing events can be validated. Comparison of Ψ values across the differentiation series revealed that hundreds of alternative exons in erythroid transcripts exhibit substantial differences in splicing efficiency between proerythroblasts and orthrochromatic erythroblasts (ΔΨ>20%), suggesting that their splicing efficiency is regulated. Both increases and decreases in exon splicing efficiency were observed, indicating that multiple splicing regulatory pathways are active and that both splicing enhancer and splicing silencer factors are involved in the regulation. Interestingly, some of the splicing switches introduce premature translation termination codons, leading us to hypothesize that splicing-coupled nonsense mediated decay may down-regulate expression of a class of erythroid transcripts. To begin exploring mechanisms that regulate the late erythroid alternative splicing program, we used the RNA-seq data to derive differentiation stage-specific expression profiles of known splicing factors. Major changes in the expression profile of many splicing regulators were observed. hnRNPA1 was strongly down-regulated in late erythroblasts, in concert with up-regulation of the protein 4.1R splicing switch it has been shown to inhibit. Because many other exons are up-regulated with similar kinetics, hnRNPA1 may be a general inhibitor of alternative splicing in early erythroblasts. In contrast, RNA-seq data indicate that several other splicing factors including MBNL1, a known splicing factor in muscle and brain whose activity is disturbed in myotonic dystrophy, are substantially up-regulated in late erythropoiesis. We conclude that a highly dynamic alternative splicing program in terminally differentiating erythroblasts, in conjunction with the better studied transcriptional program, plays a major role in regulating gene expression to insure synthesis of the appropriate constellation of proteins both quantitatively and qualitatively as the cells are remodeled in preparation for production of mature red cells.
Disclosures:
No relevant conflicts of interest to declare.
AS-Quant: Detection and Visualization of Alternative Splicing Events with RNA-seq Data
