Computational predictions of thermostability and binding affinity changes in enzymes

酶用于化学合成和再循环的使用通常受到其低热稳定性的限制，因为许多反应需要高温。因此，寻找增加蛋白质稳定性的突变的高性能计算程序将代表绿色产业的竞争优势。本计画旨在利用分子动力学模拟来确定点突变对不同酵素热稳定性的影响。这些靶标包括来自文献（RNAse SA）的用于基准测试的充分表征的实例，以及目前正在CEI实验评估的塑料降解酶的新突变体（例如，PETase和角质酶）。在折叠和未折叠状态下模拟突变以确定相对自由能变化Δ G。折叠状态基于酶的晶体结构建模，而未折叠状态可以通过水中的小肽建模。理想情况下，模拟将提供一个尺度，根据它们对蛋白质稳定性的影响对所有典型氨基酸进行排名。这可以用来训练快速的生物信息学方法。每种氨基酸类型的水合自由能和构象熵的温度依赖性数据可以帮助理解嗜热生物进化过程中的突变模式。适当的酶功能的另一个重要因素是在高温下结合底物的能力。蛋白质-配体结合通常由疏水效应驱动，这取决于温度。因此，通常需要优化结合口袋以允许在高温下（例如，PET塑料与PETase的结合）。这样的计算在计算药物设计中已经很常见，因此所需的工具已经存在于计算化学中，并且只需要修改以优化结合口袋而不是配体。可能的应用包括设计热稳定酶，可以降解塑料回收。

The use of enzymes for chemical synthesis and recycling is often limited by their low thermostability, as many reactions require high temperatures. Therefore, high performance computing procedures to find mutations that increase protein stability would represent a competitive advantage for the green industry. This project aims at using molecular dynamics simulations to determine the effect of point mutations on the thermostability of different enzymes. The targets include well-characterized examples from the literature (RNAse SA) for benchmarking, and new mutants of plastic-degrading enzymes that are currently being experimentally evaluated at the CEI (e.g., PETase and cutinases). The mutations are simulated in the folded and the unfolded state to determine the relative free energy changes deltadeltaG. The folded state is modelled based on the crystal structure of the enzyme, and the unfolded state can be modelled by small peptides in water. Ideally, the simulations will provide a scale that ranks all canonical amino acids in terms of their influence on protein stability. This can be used to train fast bioinformatics approaches. The temperature-dependent data of hydration free energies and conformational entropies of each amino acid type can help to understand the mutation patterns during the evolution of thermophilic organisms. Another important factor for proper enzyme function is the capability to bind to the substrate at elevated temperatures. Protein-ligand binding is often driven by the hydrophobic effect, which depends on the temperature. Therefore, it is often necessary to optimize the binding pocket to allow substrate binding at high temperatures (e.g., the binding of PET plastics to PETase). Such calculations are already common in computational drug design, so the required tools already exist in computational chemistry, and only have to be modified to optimize the binding pocket instead of the ligand. Possible applications include the design of thermostable enzymes that can degrade plastics for recycling.