Computational reconstruction of the LDL-receptor-related protein 1 (LRP1) atomistic structure evolution from super-tertiary to quaternary

Abstract

The blood-brain barrier is a highly complex physiological barrier that separates the blood from the central nervous system to maintain the latter’s biological equilibrium. LDL receptor-related protein 1 (LRP1) is a receptor involved in BBB transcytosis and can be used by physiological or artificially induced processes. LRP1 is critical for the trafficking of misfolded proteins such as amyloid b, hyper-phosphorylated tau, and a-synuclein. Understanding its structure and function is essential to fully understanding neurological diseases like Alzheimer’s, Parkinson’s, Huntington’s, and other related dementias. LRP1 is a modular membrane protein composed of 4544 amino acids, around 1200 of which are involved in three long and flexible structures that contain coordinated calcium ions and are decorated with small sugar chains called glycans. These three flexible components are believed to have an active role in ligand binding activity and activate a peculiar and very efficient transport mechanism. No crystal structure of LRP1 is currently available. Here, we present two LRP1 conformers representing the extremes of a conformational spectrum ranging from a completely stretched, oligomer-like structure to a stable dimeric structure. The first is based on physical and biological considerations and has been built with the RoseTTAFold deep learning tool; the second conformer is obtained via homology modelling using the experimental observation of a homologous protein, LRP2, as template. We aim to assess the two conformers’ free energy variation considering the dynamics of the flexible domains obtained with atomistic molecular dynamics while rigidly fixing b-propellers’ coordinates. The obtained conformations are solvated with the BF method to estimate the species’ free energy of solvation.