# Human FUS is toxic via association with RNA polymerase II in Drosophila

## TL;DR

This study shows that human FUS protein causes toxicity in fruit flies by interacting with RNA polymerase II, suggesting a new mechanism for neurodegenerative diseases like FTLD.

## Contribution

The study reveals a novel nuclear toxicity mechanism of FUS via interaction with RNA polymerase II, relevant to FTLD pathogenesis.

## Key findings

- Removing the nuclear localization sequence of FUS prevents toxicity in Drosophila despite cytoplasmic mislocalization.
- FUS interacts with the C-terminal domain of RNA polymerase II to induce toxicity in flies.
- Cytoplasmic mislocalization of POLR2A is observed in FUS-positive FTLD neurons but not in ALS-FUS cases.

## Abstract

The RNA-binding protein FUS is commonly mutated in familial cases of amyotrophic lateral sclerosis (ALS-FUS), where it forms cytoplasmic inclusions. In addition, non-mutated FUS is a constituent component of protein inclusions in approximately 5–10% of cases of frontotemporal lobar degeneration (FTLD). Overexpression of wild-type human FUS is toxic to Drosophila neurons, preventing normal development and shortening lifespan in adults. In this study, we demonstrated that removal of the nuclear localisation sequence (NLS) of FUS, a common consequence of ALS-associated mutations, unexpectedly prevents toxicity in Drosophila models despite inducing FUS cytoplasmic mislocalisation. Using novel flies capable of expressing mGFP-tagged FUS, we found that FUS forms dynamic protein granules in Drosophila nuclei and does not form insoluble aggregates. FUS and other FET-family paralogues interact with the repetitive disordered C-terminal domain (CTD) of the large subunit of RNA polymerase II (Polr2A). Using flies that have variable CTD repeat lengths, we demonstrated that FUS genetically interacts with the Polr2A CTD to induce toxicity. Finally, we demonstrated that this association with Polr2A could be relevant to human disease, finding that inclusion-bearing neurons of individuals with FUS-positive FTLD, but not ALS-FUS, show cytoplasmic mislocalisation of POLR2A (the Polr2A human orthologue). Together, these results imply that FUS can have a nuclear mechanism of toxicity when overexpressed in animal models. This toxicity occurs via interaction with RNA polymerase II and aberrant interaction between FUS and POLR2A may be involved in the pathogenesis of FTLD.