Toxic PR Poly‐Dipeptides Encoded by the C9orf72 Repeat Expansion Target LC Domain Polymers

Yi Lin
DOI: https://doi.org/10.1096/fasebj.31.1_supplement.760.4
2017-01-01
The FASEB Journal
Abstract:The most prevalent form of familial amyotrophic lateral sclerosis (ALS) involves an expansion of the GGGGCC (G 4 C 2 ) hexanucleotide repeat located within the first intron of a gene designated C9orf72. While normal individuals typically carry 20–30 G 4 C 2 repeats, affected patients exhibit up to 1,000 or more of the hexanucleotide repeats. Surprisingly, the extended G 4 C 2 repeats can be both transcribed and translated. Translation of either the sense or antisense transcripts of the repeats generates five poly‐dipeptides, GA n , GP n , GR n , PA n and PR n . Two of these polydipeptides, GR n and PR n , show significant cellular toxicity. Missing to date are unbiased studies of the direct intracellular targets of GR n and PR n that may explain the toxicity of these poly‐dipeptides. To address this question, we employed two independent methods, photo‐crosslinking and immunoprecipitation, to identify protein targets of PR n . Both approaches showed that the PR n targets are significantly enriched for proteins bearing low complexity (LC) sequences. By in vitro reconstitution of the interaction between PR n and hnRNP A1, we found that the LC domain of hnRNP A1 is both necessary and sufficient for the interaction with PR 20 peptide. When incubated at neutral pH and physiologic ionic strength, the LC domains of numerous RNA binding proteins polymerize into labile, cross‐β fibers. These LC domain polymers then proceed to form hydrogels that mimic the formation of RNA granules. Several independent assays give evidence that PR n binding requires that the LC domains of target proteins exist in a cross‐β polymeric state. First, a series of aliphatic alcohols (1,6‐hexanediol, 2,5‐hexanediol, etc ) showed the same hierarchical order of specificity in melting cross‐β fibers, hydrogels as they did in disrupting the interaction between the hnRNPA2 low complexity domain and the PR n poly‐dipeptide. Second, 1,6‐hexanediol, the aliphatic alcohol that has the strongest effect in dissolving cross‐β fibers, was found to impede binding of the PR n poly‐dipeptide to its native, intracellular targets. Finally, a single amino acid substitution (D290V) in hnRNP A2, which strengthens cross‐β polymer formation, also strengthens its binding to the PR n poly‐dipeptide. By constrast, another single amino acid substitution (F291S), which significantly compromises polymer formation, also impedes hnRNPA2 binding to the PR n poly‐dipeptide. In summary, our observations reveal that the intracellular targets of PR n are proteins with low complexity domains and offer mechanistic insight into the toxicity of poly‐dipeptides.
What problem does this paper attempt to address?