Sequence-encoded conformational biases shape self-assembly modes of intrinsically disordered proteins

Abstract

Self-assembly of intrinsically disordered proteins (IDPs) underlies cellular functions and disease pathogenesis. This process is mediated by two intermolecular interaction modes: transient point-to-point contacts described by the sticker-and-spacer framework, and persistent surface-to-surface contacts proposed in the cross-{beta} hypothesis. We investigated the molecular basis of these modes in the context of conformational biases, defined as sequence-encoded structural preferences of local segments. Using a five-residue model, we generated lag-series IDPs from the TIA-1 prion-like domain by systematically modulating conformational biases while preserving amino acid composition. The lag-series IDPs demonstrated distinct condensate properties and varying capacities for amyloid fibril formation. Their structural analyses revealed that strongly biased regions preferentially adopt extended structures, including {beta}-strands, and the spacing between these regions influences metastable {beta}-sheet formation. Our findings demonstrate that local conformational biases shape interaction modes of IDPs, thereby linking sequence to condensate properties and amyloid fibril formation. Significance StatementProteins fold into three-dimensional structures defined by their amino acid sequences. In contrast, intrinsically disordered proteins (IDPs) lack stable structures, yet their sequences encode unique self-assembly behaviors, including phase separation and amyloid fibril formation. Can such behaviors be explained within structure-based frameworks? Using a five-residue model, we showed that IDP sequences encode local structural preferences, termed conformational biases, within short segments. These biases determine whether segments engage in transient point-to-point interactions driving phase separation or persistent surface-to-surface interactions leading to amyloid fibril formation. By bridging sequence, structure, and interaction modes, our work provides a comprehensive mechanism for self-assembly and conceptual tools for understanding IDP-related biological functions and disease mechanisms.