Two oppositely-charged sf3b1 mutations cause defective development, impaired immune response, and aberrant selection of intronic branch sites in Drosophila.
Bei ZhangZhan DingLiang LiLing-Kun XieYu-Jie FanYong-Zhen XuPublished in: PLoS genetics (2021)
SF3B1 mutations occur in many cancers, and the highly conserved His662 residue is one of the hotspot mutation sites. To address effects on splicing and development, we constructed strains carrying point mutations at the corresponding residue His698 in Drosophila using the CRISPR-Cas9 technique. Two mutations, H698D and H698R, were selected due to their frequent presence in patients and notable opposite charges. Both the sf3b1-H698D and-H698R mutant flies exhibit developmental defects, including less egg-laying, decreased hatching rates, delayed morphogenesis and shorter lifespans. Interestingly, the H698D mutant has decreased resistance to fungal infection, while the H698R mutant shows impaired climbing ability. Consistent with these phenotypes, further analysis of RNA-seq data finds altered expression of immune response genes and changed alternative splicing of muscle and neural-related genes in the two mutants, respectively. Expression of Mef2-RB, an isoform of Mef2 gene that was downregulated due to splicing changes caused by H698R, partly rescues the climbing defects of the sf3b1-H698R mutant. Lariat sequencing reveals that the two sf3b1-H698 mutations cause aberrant selection of multiple intronic branch sites, with the H698R mutant using far upstream branch sites in the changed alternative splicing events. This study provides in vivo evidence from Drosophila that elucidates how these SF3B1 hotspot mutations alter splicing and their consequences in development and in the immune system.
Keyphrases
- immune response
- rna seq
- wild type
- crispr cas
- single cell
- poor prognosis
- genome wide
- ejection fraction
- newly diagnosed
- dendritic cells
- genome editing
- gene expression
- dna methylation
- machine learning
- prognostic factors
- transcription factor
- binding protein
- toll like receptor
- mouse model
- young adults
- deep learning
- patient reported