Back

Disulfide engineering reveals unexpected pro- and anti-aggregation conformers of human α-synuclein

Uddin, A.; Serebryany, E.

2026-01-08 biophysics
10.64898/2026.01.07.698305 bioRxiv
Show abstract

Intrinsically disordered proteins can aggregate in many distinct conformations (polymorphs). Polymorphs are a striking example of fold-switching: one primary structure able to form distinct tertiary structures. Distinct polymorphs can yield distinct molecular, cellular, and disease phenotypes. Disulfide crosslinks canalize disordered proteins into distinct regions of the conformational landscape, even if the monomer remains disordered. Human -synuclein is a disordered, natively Cys-free protein involved in synaptic transmission. Its amyloid aggregation is implicated in Parkinsons disease and other synucleinopathies. Recent cryo-EM fibril structures have revealed distinct classes of amyloid polymorphs that wild-type human -synuclein can adopt. We created three double-Cys -synuclein variants predicted to form intramolecular disulfide bridges compatible with known amyloid polymorphs, plus three other double-Cys variants at arbitrary positions as controls. We purified all six variants as intramolecularly disulfide-crosslinked monomers in solution. Two of the three predicted pro-aggregation variants formed amyloid fibrils as expected, and two of the three control variants did not. Surprisingly, one of the controls formed amyloid fibrils with a distinctive morphology by negative-stain TEM, suggesting the possibility of a previously unknown polymorph. Conversely, the crosslinked variant that failed to aggregate as expected showed sub-stoichiometric, dose-dependent anti-aggregation activity, strongly suppressing amyloid formation by wild-type -synuclein.

Matching journals

The top 6 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.