The relationship between miRNA biogenesis and RNA splicing

MiRNAs (miRNAs) are important for the proper regulation of translation, with aberrant miRNA expression contributing to diseases such as cancer. MiRNAs are expressed from longer primary-miRNA transcripts and found in a variety of genomic locations, including introns and exons of coding or long noncoding RNAs. MiRNA biogenesis begins with excision of the precursor miRNA hairpin by the Microprocessor complex which consists of two proteins: Drosha and DGCR8. Both Microprocessor cleavage and splicing occur co-transcriptionally. Splicing and Microprocessing coexist when a miRNA hairpin is located within an intron without detriment to mature mRNA or miRNA production, but little is known about how these two processes interact when a miRNA hairpin is located within an exon. Similarly, little is known about how a miRNA is processed from a long non-coding (lnc)RNA, which is an RNA transcript longer than 200 nucleotides that does not code for a protein. Intronic miRNAs are processed without detriment to the splicing of the host transcript and production of functional mRNA and protein; however, exonic miRNA processing would be expected to lead to cleavage of the exon, interfering with production of mature mRNA. To understand how the genomic location of a miRNA hairpin affects the splicing of the host gene and how splicing affects miRNA production from different locations, two different approaches were used. Firstly, liver-specific miR-122 was inserted at a series of locations within a β-globin plasmid and transfected into HeLa cells, which do not express miR-122. qPCR and northern blotting show that mature miR-122 expression is increased when located within either an intron or exon compared to its endogenous context within a long noncoding RNA. Importantly, the levels of spliced β-globin was shown to decrease when miR-122 was expressed from an exon. This suggests that there is some competition between miRNA biogenesis and splicing. Secondly, the effect of the splice inhibitor pla B was used to understand the effect of splicing on endogenous pri-miR-122. Splice inhibition led to a decrease in the levels of spliced pri-miR-122, but interestingly also led to a decrease in the levels of unspliced transcript indicating that splice inhibition strongly reduces the transcription of pri-miR-122. This was found to be unique to the endogenous gene, as it was not seen when pri-miR-122 was ectopically expressed in HeLa cells. Analysis of the chromatin-associated RNA further confirmed that splice inhibition was having a strong negative effect on transcription. Transcription of pri-miR-122 is known to be terminated by Microprocessing rather than the canonical pathway, and analysis of other lnc-pri-miRNAs which are Microprocessor terminated showed that the effect of splice inhibition on transcription was not unique to pri-miR-122, but was also shared by the important oncogene pri-miR-17~92a. The results in this thesis show that intronic or exonic miRNA hairpins do have an effect on splicing, and unexpectedly that splicing is important for efficient transcription of specific lnc-pri-miRNAs. This work shows that the relationship between Microprocessing and splicing, as well as transcription, is complex. By elucidating how these co-transcriptional processes occur on the same transcript, the understanding of how miRNA biogenesis is regulated from different genomic locations will be improved. This could further the current knowledge of how miRNA disregulation occurs in disease and may lead to the development of new drugs to treat them