TDP-43 fibrils emerge when a short charged motif crosses a charge threshold
A new preprint reframes how TDP-43 converts from a phase-separated protein into an amyloid-forming species inside cells. The central finding is stark: fibril formation tracks the net charge of a polyampholyte motif in the linker between RRM2 and the low-complexity domain, not the degree to which a condensate has solidified. In other words, condensate aging alone does not explain the transition to pathological amyloid.
Investigators worked with budding yeast expressing GFP-TDP-43ΔNLS and focused on a charge reversal mutant in which three arginines were replaced by glutamates (R268E/R272E/R275E; 3R-E). This variant changed liquid puncta into ProteoStat-positive fibrils, a readout consistent with amyloid conversion. The result points to a local electrostatic switch rather than a broad loss of protein mobility as the decisive event.
Condensate mobility and fibrillization moved in opposite directions
One of the most striking observations came from fluorescence recovery after photobleaching. The 3R-E condensates were more mobile than wild-type condensates, yet they were the ones that converted into fibrils. By contrast, the least mobile condensates in both yeast and human U2OS cells remained ProteoStat-negative. A simple hardening model does not fit those data.
This matters because TDP-43 biology has long been discussed through the lens of liquid-liquid phase separation, with the assumption that maturation from liquid droplets to denser assemblies increases amyloid risk. These experiments separate two processes that are often conflated. Condensate viscosity and fibrillization are related only indirectly here, and the decisive variable appears to be the charge state of a narrow sequence element rather than the mechanical rigidity of the puncta themselves.
The biological logic is plausible. A polyampholyte motif can alter local electrostatic balance, shifting intramolecular contacts and changing how the amyloid-prone low-complexity domain samples conformations. If charge redistribution makes the fibril-forming region more accessible or less effectively restrained, amyloid assembly could accelerate even in a condensate that remains highly dynamic.
Net charge predicted fibrils across 24 variants
The strongest evidence came from a broader mutational series. Across 24 PAM charge variants, the proportion of fibril-bearing cells rose linearly with net charge, independent of residue identity. The analysis did not depend on whether the altered residues were arginines, serines, or other positions within the motif; what mattered was the net electrostatic state.
Neutralizing the motif delayed conversion rather than abolishing it. That detail is important because it suggests the sequence operates as a tunable rheostat, not an all-or-none gate. Cells may tolerate some charge drift before crossing the threshold for amyloid assembly, but once the balance shifts far enough, fibrillization becomes increasingly likely.
The architecture of the protein explains why this could be so effective. The polyampholyte motif sits in a linker between RRM2 and the low-complexity domain, positioning it to influence both RNA-associated interactions and the amyloid-prone region that drives aggregation. A compact charge-sensing element in that location can reshape the conformational landscape without requiring large-scale structural rearrangement elsewhere in the protein.
Phosphomimetic substitution points to a regulated charge switch
The study also explored whether phosphorylation could modulate this charge-sensitive behavior. Phosphomimetic substitution of the PAM serines alone induced fibrils, implicating phosphorylation as a potential physiological regulator of the motif. That is a consequential clue, because phosphorylation would provide a reversible way for cells to alter the net charge of TDP-43 and shift its aggregation propensity.
If true in native systems, the finding offers a mechanistic bridge between signaling and proteinopathy. Kinase activity could alter the electrostatic profile of the motif, and a change in charge could push TDP-43 toward pathological self-association even before gross condensate solidification appears. The result does not claim phosphorylation is the only relevant input, but it places post-translational modification squarely inside the fibrillization mechanism.
Why the charge model changes the way TDP-43 pathology is read
TDP-43 is central to ALS and FTLD biology, and much of the field has tried to explain disease progression through condensate maturation. This preprint argues for a more specific control point. The formation of amyloid-like assemblies in cells can be predicted from a local charge code embedded in the protein sequence, while condensate fluidity alone fails to separate fibril-positive from fibril-negative states.
That distinction reshapes how one thinks about intervention. Strategies aimed only at preventing condensate hardening may miss an upstream electrostatic trigger that governs whether TDP-43 crosses into the amyloid state. A therapy that stabilizes droplets but leaves the PAM charge state permissive could still allow fibril formation.
The work also leaves open several practical questions for the field. The experiments were done in yeast and human U2OS cells, so the next step is to determine how faithfully this charge rule holds in neurons, where TDP-43 pathology is most clinically relevant. The preprint also does not establish how endogenous phosphorylation patterns map onto the motif in disease tissue, or which signaling pathways are most likely to tune that charge state in vivo.
Even so, the message from this dataset is direct. TDP-43 amyloid conversion in cells follows the electrostatic status of a polyampholyte motif, and condensate solidification is not the primary arbiter of whether fibrils appear. In a protein long discussed for its phase behavior, the local charge balance now looks like the more precise molecular switch.