Automated and statistically corrected identification of flexible multivalent IDP-bound assemblies in electron micrographs

Abstract

Multivalent intrinsically disordered protein (IDP) complexes are prevalent in biology and control diverse cellular functions, including tuning levels of transcription, coordinating cell-signaling events, and regulating the assembly and disassembly of complex macromolecular architectures. These systems pose a significant challenge to structural investigation, due to the continuum dynamics imparted by the IDP and compositional heterogeneity resulting from characteristic low-affinity interactions. Traditional single-particle electron microscopy (EM) is a powerful tool for visualizing IDP complexes. However, the IDPs themselves are typically "invisible" by EM, undermining methods of image analysis and structural interpretation. To overcome these challenges, we developed a pipeline for automated analysis of common beads-on-a-string type of assemblies, composed of IDPs bound at multivalent sites to the ubiquitous ~20 kDa cross-linking hub protein LC8. This approach quantifies conformational and compositional heterogeneity on a single-particle basis, and statistically corrects spurious observations arising from random proximity of bound and unbound LC8. After careful validation of the methodology, the approach was applied to the nuclear pore IDP Nup159 and the transcription factor ASCIZ. The analysis unveiled significant compositional and conformational diversity in both systems that could not be obtained from traditional single particle EM class-averaging strategies, and shed new light on how these architectural properties contribute to their physiological roles in supramolecular assembly and transcriptional regulation. Ultimately, we expect that this approach may be adopted to many other intrinsically disordered systems that have evaded traditional methods of structural characterization. Significance StatementIntrinsically disordered proteins (IDPs) or protein regions (IDRs) represent >30% of the human proteome, but mechanistically remain some of the most poorly understood classes of proteins in biology. This dearth in understanding stems from these very same intrinsic and dynamic properties, which make them difficult targets for quantitative and structural characterization. Here, we present an automated approach for extracting quantitative descriptions of conformational and compositional heterogeneity present in a common beads-on-a-string type of multivalent IDP system from single-particle images in electron micrographs. This promising approach may be adopted to many other intrinsically disordered systems that have evaded traditional ensemble methods of characterization.