Back

C9orf72-linked arginine-rich dipeptide repeats aggravate pathological phase separation of G3BP1

Nerom, M. V.; Ahmed, J.; Lazar, T.; Lindt, J. V.; Pancsa, R.; Maes, D.; Tompa, P.

2023-03-31 molecular biology
10.1101/2023.03.31.535023 bioRxiv
Show abstract

Ras GTPase-activating protein-binding protein 1 (G3BP1) is the key protein driving the formation of cytoplasmic stress granules (SGs) by liquid-liquid phase separation (LLPS). It is a switch-like protein held in a closed and inactive state by intramolecular electrostatic interactions competitively opened by RNA, activating the protein and initiating its LLPS. Here we show that C9orf72-derived arginine-rich dipeptide repeats PR30 and GR30 (R-DPRs) present in amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD), also bind to G3BP1, switching it to an LLPS-competent open state much more effectively than RNA. Whereas RNA binds G3BP1 with micromolar affinity, and cannot initiate LLPS without crowding agents, R-DPRs exhibit a thousand-fold stronger binding to G3BP1, eliciting rapid LLPS even without crowding. The pathogenic effect of R-DPRs is also underscored by the slow transition of R-DPR-G3BP1 liquid droplets to aggregated, ThS-positive states that can recruit the ALS-linked protein hnRNPA2. Deletion constructs and molecular simulations show that R-DPR binding and LLPS are mediated via binding through the negatively charged intrinsically disordered region 1 (IDR1) of the protein, allosterically regulated by the positively charged IDR3. Bioinformatic analyses point to the strong mechanistic parallels of these effects with the interaction of R-DPRs with nuclear nucleophosmin (NPM1) and also suggest that R-DPRs also interact with many other similar nucleolar and stress-granule proteins, extending the underlying mechanism of R-DPR toxicity in cells.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.