Seeding the aggregation of TDP-43 requires post-fibrillization proteolytic cleavage

Abstract

Despite the strong evidence linking the transactive response DNA-binding protein 43 (TDP-43) aggregation to the pathogenesis of frontotemporal lobar degeneration with TDP-43, amyotrophic lateral sclerosis and several neurodegenerative diseases, our knowledge of the sequence and structural determinants of its aggregation and neurotoxicity remains incomplete. Herein, we present a new method for producing recombinant full-length TDP-43 filaments that exhibit sequence and morphological features similar to those of brain-derived TDP-43 filaments. We show that TDP-43 filaments contain a β-sheet-rich helical amyloid core that is fully buried by the flanking structured domains of the protein. We demonstrate that the proteolytic cleavage of TDP-43 filaments and exposure of this amyloid core are necessary for propagating TDP-43 pathology and enhancing the seeding of brain-derived TDP-43 aggregates. Only TDP-43 filaments with exposed amyloid core efficiently seeded the aggregation of endogenous TDP-43 in cells. These findings suggest that inhibiting the enzymes mediating cleavage of TDP-43 aggregates represents a viable disease-modifying strategy to slow the progression of amyotrophic lateral sclerosis and other TDP-43 proteinopathies.

Recombinant Full-Length TDP-43 Oligomers Retain Their Ability to Bind RNAs, Are Not Toxic, and Do Not Seed TDP-43 Aggregation in Vitro

“…Several studies have suggested that the oligomeric forms of several amyloid proteins play central roles in mediating amyloid toxicity and cell-to-cell propagation of protein aggregates; this process is strongly dependent on their seeding activity, i.e., the ability to induce protein misfolding and fibril formation. − Therefore, we next investigated the seeding activity of TDP-43 oligomers. The full-length TDP-43 aggregates and fibrils do not bind to thioflavin T (ThT) because their amyloid core is buried and covered by the RNA-binding domains; thus, we investigated the ability of TDP-43 oligomers to seed the aggregation-prone core peptide corresponding to residues 279–360 as described previously . This TDP-43 fragment forms highly ordered fibrils that bind to ThT, and its aggregation is accelerated by the addition of TDP-43 (279–360) fibrillar seeds (Figure C,D and Figure S2).…”

“…The N-terminal His-SUMO fusion protein was prepared following a protocol previously developed in our laboratory. 40 Briefly, the plasmid was transformed and overexpressed in E. coli strain BER2566. A 3 L bacterial culture was harvested at 4000 rpm for 15 min at 4 °C.…”

“…The peptide was prepared according to our recent paper. 21 Then, 150 μg of lyophilized TDP-43 core peptide 279−360 was disaggregated in 100 μL of HFIP/TFA (1:1) buffer at 30 °C for 30 min. The solution was evaporated under a mild nitrogen stream.…”

“…Previously, our laboratory described an efficient protocol to produce full-length TDP-43 in E. coli. 21 Briefly, TDP-43 was firstly expressed as a NTD-tagged His-SUMO fusion protein and purified by a HisTrap column. After the His-SUMO fusion tag was removed using ubiquitin-like-specific protease 1 (Ulp-1) overnight, TDP-43 rapidly formed oligomers that could be separated from monomeric TDP-43 by SEC (Figure 1A).…”

“…Proteinase K (ProK)-mediated proteolytic cleavage of the full-length fibrils or brain-derived fibrils leads to exposure of the amyloid core and converts them into seeding competent fibrils; i.e., they induce seeding of TDP-43 (279−360). 40 Therefore, the lack of seeding activity by the oligomers could also be due to the inaccessibility of the amyloid core, which was buried by the flanking RNA binding domains.…”