Abstract
Nuclear transport relies on intrinsically disordered FG-nucleoporins (FG-Nups) that form a dynamic selective barrier, yet experiments show that these proteins can also adopt highly ordered fibrillar structures. Capturing this duality within a single computational framework remains a major challenge. Here, we introduce 2BPA-HB, a sequence-resolved coarse-grained model that combines explicit, directional backbone hydrogen bonding with residue-specific side chain interaction centers derived from all-atom simulations. This design enables microsecond-scale simulations of the transitions between disordered, condensed, and amyloid-like states while maintaining computational efficiency. We first validate the model by simulating several experimentally resolved Nup98 fibril polymorphs, which retain their characteristic β-sheet registry and protofibril architecture. Seeded growth simulations further show that disordered FG fragments align with and extend these fibrillar templates in a sequence-specific manner, and can initiate secondary nucleation events on fibril surfaces. Applying the model to yeast FG-Nups, we find that cohesive FG domains form stable condensates with interaction networks dominated by FG motif contacts, consistent with experimental observations and prior simulations. The increased structural resolution of 2BPA-HB additionally reveals transient α- and β-structures within these condensates. Together, these results demonstrate that 2BPA-HB captures both liquid-like and fibrillar behavior of intrinsically disordered proteins within a single coarse-grained framework. The model provides a platform to study how disorder, secondary structure formation, and aggregation coexist in nuclear pore biology, and represents a step forward to a transferable coarse-grained approach for low-complexity proteins that undergo both phase separation and amyloid formation, as implicated in neurodegenerative diseases.
Full text
2,020 characters
· extracted from
oa-doi-fallback
· click to expand
Abstract
Nuclear transport relies on intrinsically disordered FG-nucleoporins (FG-Nups) that form a dynamic selective barrier, yet experiments show that these proteins can also adopt highly ordered fibrillar structures. Capturing this duality within a single computational framework remains a major challenge. Here, we introduce 2BPA-HB, a sequence-resolved coarse-grained model that combines explicit, directional backbone hydrogen bonding with residue-specific side chain interaction centers derived from all-atom simulations. This design enables microsecond-scale simulations of the transitions between disordered, condensed, and amyloid-like states while maintaining computational efficiency. We first validate the model by simulating several experimentally resolved Nup98 fibril polymorphs, which retain their characteristic β-sheet registry and protofibril architecture. Seeded growth simulations further show that disordered FG fragments align with and extend these fibrillar templates in a sequence-specific manner, and can initiate secondary nucleation events on fibril surfaces. Applying the model to yeast FG-Nups, we find that cohesive FG domains form stable condensates with interaction networks dominated by FG motif contacts, consistent with experimental observations and prior simulations. The increased structural resolution of 2BPA-HB additionally reveals transient α- and β-structures within these condensates. Together, these results demonstrate that 2BPA-HB captures both liquid-like and fibrillar behavior of intrinsically disordered proteins within a single coarse-grained framework. The model provides a platform to study how disorder, secondary structure formation, and aggregation coexist in nuclear pore biology, and represents a step forward to a transferable coarse-grained approach for low-complexity proteins that undergo both phase separation and amyloid formation, as implicated in neurodegenerative diseases.
Competing Interest Statement
The authors have declared no competing interest.
Text is read by the "Ask this paper" AI Q&A widget below.
Extraction quality varies by source — PMC NXML preserves structure
cleanly, OA-HTML may include some navigation residue, and OA-PDF can
have broken hyphenation. The publisher copy
(via DOI)
is the canonical version.