Combining AlphaFold and Molecular Dynamics Simulations to Predict the Complete Structures of Isoforms of Human Nerve Growth Factor and Human Brain-Derived Neurotrophic Factor

Abstract

Neurotrophins play an essential role in the development and maintenance of the central and peripheral nervous systems. There are four members of the mammalian neurotrophin family of growth factors, and they include nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3), and neurotrophin-4/5 (NT-4/5). NGF and BDNF have garnered much attention given that they are found in significant concentrations in the brain, and are associated with learning, memory, aging, psychiatric conditions, and neurodegenerative diseases. There are multiple physiologically relevant isoforms of NGF (mature NGF, proNGF-B, proNGF-A) and BDNF (mature BDNF, truncated BDNF, proBDNF), and the mechanisms through which they determine neuronal signalling events are complex and not fully understood. Given the limited information available from resolved structures, here we combined AlphaFold and Molecular Dynamics simulations to determine the similarities and differences in conformation across all physiologically relevant isoforms of NGF and BDNF, respectively. Structural and dynamic differences were measured between the various segments of the precursor isoforms of NGF and BDNF, as the N-terminal disordered regions transitioned from an extended to a compact conformation. Moreover, global conformational differences were observed between the proneurotrophin and the corresponding mature neurotrophin isoforms, in part due to a possible V-shaped versus parallel arrangement of the core mature domain. Additionally, the added use of MD simulations provided insights that were not apparent from observing the AlphaFold-generated structures only. These included differences between both chains of the homodimer, structural deviations when comparing the post-simulation conformations to the structures in the Protein Data Bank, and new conformations of the neurotrophins due to shifts in the positioning of structural features. Ultimately, the use of computational methods revealed structural similarities and differences when comparing across the isoforms of NGF and BDNF, respectively, while also more generally finding that MD simulations can add value when combined with AlphaFold.