Deep manifold learning reveals hidden dynamics of proteasome autoregulation

• DOI: 10.48550/arxiv.2012.12854
• Publication Date: 2020-01-01
• Authors: Zhaolong Wu; Shuwen Zhang; Wei Li Wang; Yinping Ma; Yuanchen Dong; Youdong Mao
• Keywords: Proteostasis

The 2.5-MDa 26S proteasome maintains proteostasis and regulates myriad cellular processes. How polyubiquitylated substrate interactions regulate activity is not understood. Here we introduce a deep manifold learning framework, named AlphaCryo4D, which enables atomic-level cryogenic electron microscopy (cryo-EM) reconstructions of nonequilibrium conformational continuum reconstitutes hidden dynamics autoregulation in the act degradation. AlphaCryo4D integrates 3D residual with embedding free-energy landscapes, directs clustering via an energy-based particle-voting algorithm. In blind assessments using simulated heterogeneous cryo-EM datasets, achieved classification accuracy three times that conventional method reconstructed continuous changes 130-kDa protein at sub-3-angstrom resolution. By to analyze single experimental dataset, identified 64 conformers substrate-bound human proteasome, revealing entanglement two regulatory particles doubly capped holoenzymes their energetic differences singly ones. Novel ubiquitin-binding sites are discovered on RPN2, RPN10 Alpha5 subunits remodel polyubiquitin chains for deubiquitylation recycle. Importantly, choreographs single-nucleotide-exchange proteasomal AAA-ATPase motor during translocation initiation, upregulates proteolytic by allosterically promoting nucleophilic attack. Our systemic analysis illuminates grand hierarchical allostery autoregulation.