Processive and Distributive Non-Equilibrium Networks Discriminate in Alternate Limits

• Published in: arXiv.org 2022-06
• Main Authors: Venkataraman, Gaurav G, Miska, Eric A, Jordan, David J
• Format: Article
• Language: English
• Subjects: Amplification; Data processing; Discrimination; Editing; Energy dissipation; Equilibrium; Liquid phases; Orthogonality; Sensitivity; Steady state
• ISSN: 2331-8422
• DOI: 10.48550/arxiv.2001.11918

We study biochemical reaction networks capable of product discrimination inspired by biological proofreading mechanisms. At equilibrium, product discrimination, the selective formation of a "correct" product with respect to an "incorrect product", is fundamentally limited by the free energy difference between the two products. However, biological systems often far exceed this limit, by using discriminatory networks that expend free energy to maintain non-equilibrium steady states. Non-equilibrium systems are notoriously difficult to analyze and no systematic methods exist for determining parameter regimes which maximize discrimination. Here we introduce a measure that can be computed directly from the biochemical rate constants and which provides a necessary condition for proofreading. Non-equilibrium systems can discriminate using binding energy (energetic) differences or using activation energy (kinetic) differences, but in both cases, proofreading is optimal when dissipation is maximized, so long as the necessary condition we introduce is satisfied. We show that these constraints can be in opposition and that this explains why the error rate is a non-monotonic function of the irreversible drive in the original proofreading scheme of Hopfield and Ninio. Finally, we introduce mixed networks, in which one product is favored energetically and the other kinetically. In such networks, sensitive product switching can be achieved simply by spending free energy. Biologically, this corresponds to the ability to select between products by driving a single reaction without network fine tuning. This may be used to explore alternate product spaces in challenging environments.