Constructing catalytically proficient enzymes from de novo designed proteins

酶是基本上重要的生物分子，在所有生物体中执行大部分化学反应。它们本身就是蛋白质，由氨基酸链组成，尽管它们与蛋白质的区别在于它们能够大大增加化学反应的速率。这些反应为细胞生命提供动力，并参与了大量赋予细胞化学和物理特性的基本过程。许多酶进行具有实质性商业或医学价值的化学反应，因为转化的产物可以是药物、燃料或其他有用的物质或材料。对于这样重要或有用的反应，通常没有人造物质可以像酶那样精确或有效地催化特定的化学转化。还有许多化学转化尚未发现酶。因此，人们对构建能够进行选定化学反应的定制酶有着巨大的兴趣。虽然在过去的100年里，我们对天然酶催化作用有了令人难以置信的强大理解，但人工设计的酶和自然界中发现的酶的能力仍然存在不足。我们认为，这在一定程度上是由于普遍使用自然进化的蛋白质作为创造人工酶的起点。这些天然蛋白质可能是脆弱的，难以纯化且昂贵，在细胞环境中无活性，对有机分子和溶剂化学敏感，最重要的是，它们总是带来进化的复杂性，这可能会阻碍酶设计者的修饰。我们相信，这种进化的包袱不是蛋白质和酶的必要特征，在某些情况下，最好是与不受自然选择影响的蛋白质一起工作。我们的简单蛋白质，称为maquettes，是不含天然蛋白质序列的小而坚固的蛋白质支架，因此不受进化带来的任何复杂性的影响。通常情况下，我们设计这些模型，包括一个非蛋白质分子，赋予自己的反应性的支架。血红素分子是一种特别通用的分子，我们在设计中包括了它，它存在于大量的天然酶中，其中许多酶催化非常具有挑战性的化学反应。我们已经使用了基本的设计过程来将含血红素的模型开发成活性人工酶，其功能与许多天然酶一样用于去除电子-氧化-从各种各样的基板。它甚至可以比专门为此目的进化的天然酶更高效地对常见污染物进行解毒。人工酶对温度和有机溶剂的存在相对不敏感，并且是设计在工业生物技术中具有巨大潜力的新型、廉价和高效生物催化剂的极好起点。我们在这里提出的工作旨在利用这一最近的成功，并开发各种各样的maquettes，这些maquettes将作为强大的人工酶，能够催化几种具有商业价值和挑战性的反应。通过了解天然酶的结构和功能，我们将使用强大的新计算方法以及迭代的实验方法来实现这一目标。重要的是，这些反应包括在自然界中未观察到的反应，由此产生的产物含有不寻常的和高度紧张的环结构，并具有显著的生物活性（例如药物，杀虫剂）。由于模型在细菌中完全和功能性地组装，我们还可以采用强大的高通量实验室进化策略以半随机方式提高催化活性。

Enzymes are fundamentally important biological molecules that perform the bulk of the chemical reactions in all living organisms. The are themselves proteins, made up of chains of amino acids, though what differentiates them from proteins is their ability to massively increase the rate of chemical reactions. These reactions power cellular life and are involved in a great number of essential processes that give cells their chemical and physical characteristics. Many enzymes perform chemical reactions which have substantial commercial or medical value, as the products of the transformations may be drugs, fuels or other useful substances or materials. It is often the case that for such important or useful reactions, there are no manmade substances available to catalyse the specific chemical transformations with the same degree of precision or efficiency as enzymes. There are also many chemical transformations for which no enzyme has yet been discovered. Therefore, there is a huge interest in building tailor-made enzymes capable of performing selected chemical reactions. While we have gained an incredibly powerful understanding of natural enzymatic catalysis over the past 100 years, there remains a shortfall in the capabilities of artificial, designed enzymes and those found in nature. We believe that this due, in part, to the prevalent use of naturally evolved proteins as the starting points for creating artificial enzymes. These natural proteins may be fragile, difficult and expensive to purify, inactive out of their cellular environment, chemically sensitive to organic molecules and solvents, and, most significantly, they invariably bring an evolutionary complexity with them that can hinder modification by the enzyme designer. We believe that this evolutionary baggage is not a necessary feature of proteins and enzymes and that in certain cases, it might be preferable to work with proteins untouched by natural selection. Our simple proteins, called maquettes, are small robust protein scaffolds that contain no natural protein sequences, and are therefore free from any complexity imposed by evolution. Typically, we design these maquettes to include a non-protein molecule that imparts its own reactivity onto the scaffold. The heme molecule is a particularly versatile molecule that we include in our designs, and it is present in a plethora of natural enzymes, many of which catalyse exceptionally challenging chemical reactions.With our most recent work, we have used an elementary design process to develop a heme-containing maquette into an active artificial enzyme that functions as well as many natural enzymes for the removal of electrons - oxidation - from a broad range of substrates. It can even perform the detoxification of a common pollutant with higher efficiency than a natural enzyme that has evolved specifically for this purpose. The artificial enzyme is relatively insensitive towards temperature and the presence of organic solvents, and is an excellent starting point for the design of new, cheap and highly efficient biocatalysts that have huge potential in industrial biotechnology. The work we propose here aims to exploit this recent success and develop a diverse range of maquettes that will act as robust artificial enzymes capable of catalysing several commercially valuable and challenging reactions. Informed by the structures and our functional understanding of natural enzymes, we will use powerful new computational methods alongside iterative, experimental approaches to achieve this. Crucially, these include reactions not observed in nature, whereby the resulting products contain unusual and highly strained ring structures, and have significant biological activities (e.g. drugs, insecticides). Since the maquettes are fully and functionally assembled in bacteria, we can also employ powerful, high throughput laboratory evolution strategies to improve catalytically activity in a semi random manner.