Post-translational modifications (PTMs) have emerged as key modulators of protein phase separation and have been linked to protein aggregation in neurodegenerative disorders. The major aggregating protein in amyotrophic lateral sclerosis and frontotemporal dementia, the RNA-binding protein TAR DNA-binding protein (TDP-43), is hyperphosphorylated in disease on several Cterminal serine residues, a process generally believed to promote TDP-43 aggregation. Here, we however find that Casein kinase 1δmediated TDP-43 hyperphosphorylation or C-terminal phosphomimetic mutations reduce TDP-43 phase separation and aggregation, and instead render TDP-43 condensates more liquid-like and dynamic. Multi-scale molecular dynamics simulations reveal reduced homotypic interactions of TDP-43 low-complexity domains through enhanced solvation of phosphomimetic residues. Cellular experiments show that phosphomimetic substitutions do not affect nuclear import or RNA regulatory functions of TDP-43, but suppress accumulation of TDP-43 in membrane-less organelles and promote its solubility in neurons. We speculate that TDP-43 hyperphosphorylation may be a protective cellular response to counteract TDP-43 aggregation.

No abstract

Cytoplasmic aggregation and concomitant nuclear clearance of the RNA-binding protein TDP-43 are found in ~ 90% of cases of amyotrophic lateral sclerosis and ~ 45% of patients living with frontotemporal lobar degeneration, but no disease-modifying therapy is available. Antibody therapy targeting other aggregating proteins associated with neurodegenerative disorders has shown beneficial effects in animal models and clinical trials. The most effective epitopes for safe antibody therapy targeting TDP-43 are unknown. Here, we identified safe and effective epitopes in TDP-43 for active and potential future passive immunotherapy. We prescreened 15 peptide antigens covering all regions of TDP-43 to identify the most immunogenic epitopes and to raise novel monoclonal antibodies in wild-type mice. Most peptides induced a considerable antibody response and no antigen triggered obvious side effects. Thus, we immunized mice with rapidly progressing TDP-43 proteinopathy (“rNLS8” model) with the nine most immunogenic peptides in five pools prior to TDP-43ΔNLS transgene induction. Strikingly, combined administration of two N-terminal peptides induced genetic background-specific sudden lethality in several mice and was therefore discontinued. Despite a strong antibody response, no TDP-43 peptide prevented the rapid body weight loss or reduced phospho-TDP-43 levels as well as the profound astrogliosis and microgliosis in rNLS8 mice. However, immunization with a C-terminal peptide containing the disease-associated phospho-serines 409/410 significantly lowered serum neurofilament light chain levels, indicative of reduced neuroaxonal damage. Transcriptomic profiling showed a pronounced neuroinflammatory signature (IL-1β, TNF-α, NfκB) in rNLS8 mice and suggested modest benefits of immunization targeting the glycine-rich region. Several novel monoclonal antibodies targeting the glycine-rich domain potently reduced phase separation and aggregation of TDP-43 in vitro and prevented cellular uptake of preformed aggregates. Our unbiased screen suggests that targeting the RRM2 domain and the C-terminal region of TDP-43 by active or passive immunization may be beneficial in TDP-43 proteinopathies by inhibiting cardinal processes of disease progression. Graphical Abstract

No abstract

No abstract