Phd in GENETICS AND MOLECULAR BIOLOGY

Thesis title: TGS1 and SMN cooperate to preserve transcriptome fidelity in human cells and flies and protect against neuronal death in a Drosophila model for SMA

Spinal muscular atrophy (SMA) is a devastating neurodegenerative disease caused by a deficiency of the Survival of Motor Neuron protein (SMN), which is crucial for the assembly of spliceosomal snRNPs. Cells depleted for SMN accumulate aberrantly spliced transcripts. In human cells, SMN cooperates with the hypermethylase TGS1, an S-adenosyl-L-methionine dependent methyltransferase responsible for snRNAs cap trimethylation. We found that TGS1 loss in HeLa cells results in the accumulation of unprocessed snRNAs, with U2 snRNA being the most affected species, carrying extended 3' tails that are often uridylated and these defects are also present in Drosophila Tgs1 mutants. In addition, loss of Tgs1 in Drosophila results in neurodegenerative phenotypes similar to those caused by Smn depletion. Analyses conducted in the Drosophila model revealed that Tgs1 physically and genetically interacts with all the subunits of Smn complex. By using an RNAi-based system to deplete Tgs1 and Smn in the fly eye imaginal discs, we have shown that both proteins are required for retinal progenitor cells viability and that the Smn loss of function phenotype is rescued by overexpression of Tgs1 and viceversa. The depletion of Tgs1 and Smn in the Drosophila neuroepithelium is a convenient system to study the mechanisms of cell death and to explore how Tgs1 and Smn cooperate to protect against neurodegeneration. Notably, expression of human TGS1 ameliorates Smn-dependent phenotypes in both flies and worms, suggesting that TGS1 acts as a conserved SMN modifier. It is believed that alterations in the maturation and processing of snRNAs could have an impact on motor neurons diseases, including SMA, by altering splicing and expression of transcripts necessary to prevent motor neurons death. We combined the power of short (Illumina) and long read (ONT) sequencing to perform extensive transcriptome profiling in human HeLa cells deficient for TGS1 and SMN. In both mutant backgrounds, we found extensive transcriptome alterations with a partial overlap in the expression of aberrantly spliced transcripts. Interestingly, a fraction of the aberrant RNA molecules carried unspliced introns and/or 3' elongated (readthrough) transcripts that often extended into the adjacent intergenic region and the downstream gene. These transcripts are likely generated by the co-occurrence of inefficient splicing and defective transcription termination. We also found that these aberrant molecules accumulate in motoneurons differentiated from iPSCs of type I SMA patients. It has been proposed that in certain human cell types, SMN may be involved in transcription termination and prevents the accumulation of DNA:RNA hybrids (R-loops) at the end of genes. We have found that loss of Tgs1 and Smn in Drosophila larval brains results in abundant R-loops and DNA damage foci that are rescued by overexpression of human RNaseH1. Our results support a working model in which the accumulation of aberrant RNAs is a source of transcriptional stress and DNA damage that induces neurotoxic effects and contributes to degeneration in neurons. By combining deep transcriptome profiling with the analysis of cell survival, this work aims at elucidating the functional correlations between defective splicing, transcription termination and R-loops and to which extent R-loop accumulation affects the survival of neurons, thus contributing to SMA pathogenesis.