Ultra-high throughput evolution of designer enzymes with extended amino acid alphabets

随着社会和政治对环境和气候问题的日益认识，各个工业部门正在将重点转向碳中和和环保技术。因此，生物催化是化学工业中用于生产商品化学品和药品的一项迅速发展的技术。在生物催化中，酶作为自然界的催化剂，在人类设计的过程中被重新用于合成化学物质。酶可以加速高度复杂的化学反应，其速度和特异性是传统化学方法无法比拟的。此外，反应可以在低温水溶液中进行，而不像化学过程通常需要高温、有毒化学物质和大量有机溶剂。然而，酶的广泛工业开发的一个主要限制是，在自然界中没有一种合适的酶可用于任何所需的反应。因此，迫切需要符合有机化学家规格的酶。而不是重新设计现有的天然酶，自下而上的新酶设计现在成为一种可行的选择。事实上，通过计算和实验优化，已经成功地创造了用于简单反应的高效人工酶。然而，可以设计的化学机制的数量固有地局限于官能团和氨基酸的自然曲目。通过遗传密码的扩展，现在有可能用额外的功能来增加自然界的氨基酸字母表。特别是，通过将氨基酸融合到小分子催化剂中，有机化学家已经对其进行了广泛的探索，现在有可能创造出具有全新反应范围的酶，这在自然界是前所未有的。为了从零开始设计具有非天然功能的更好的酶，有必要获得对人工氨基酸必须如何放置和进一步补充酶活性位点的基本理解。该项目旨在通过实验改进带有非天然氨基酸的设计酶来解决这个问题，从而对使用额外构建块增强自然的真正潜力有了基本的了解。这将通过定向进化来实现，这是在实验室时间尺度上模拟达尔文进化。定向进化允许发现改善酶活性的突变，但这些突变在理性上是不可预测的。通过反复的诱变和选择，酶的活性水平可以显著提高。这是一个艰苦的过程，因为许多酶变体必须单独分析，以确定具有有益作用的罕见突变。为了加速这一过程，将实施超高通量分析，利用皮升大小的液滴作为含有酶变体及其编码基因的反应容器。液滴可以以每秒几千个液滴的速度进行操作，最重要的是，它们可以根据酶的活性进行分类。有了这项技术，现在有可能首次探索这些设计酶在不同不相关反应中的进化极限，包括酯水解和非自然碳-碳键形成反应。深入表征性能最好的酶将突出功能和结构特征，这是支持非天然氨基酸催化所必需的。通过计算酶设计来概括这些发现将挑战我们已经获得的分子理解，并产生新一代改进的设计酶。总的来说，本研究项目将为开发用于非生物反应的高活性酶开辟新的途径，在生物技术、生物催化和合成生物学方面具有重要意义。

With increasing societal and political acknowledgment of environment and climate issues, various industrial sectors are shifting their focus towards carbon neutral and environmentally benign technologies. As thus, biocatalysis is a rapidly expanding technology in chemical industry for the production of commodity chemicals and pharmaceuticals. In biocatalysis, enzymes, which are nature's catalysts, are repurposed for synthesizing chemicals in human devised processes. Enzymes can accelerate highly complex chemical reactions with speeds and specificities that are unrivalled by conventional chemical methods. In addition, reactions can be performed at low temperatures in aqueous solutions, unlike chemical processes that typically require high temperatures, toxic chemicals and large amounts of organic solvents. However, a major limitation for broad industrial exploitation of enzymes is that not for any desired reaction a suitable enzyme is available in nature's repertoire. As thus, enzymes that meet the specifications of organic chemists are urgently required. Rather than reengineering existing natural enzymes, bottom-up design of new enzymes is now becoming a feasible alternative. In fact, highly efficient artificial enzymes for simplistic reactions were successfully created through computation and experimental optimisation. The number of chemical mechanisms that can be designed is however inherently limited to natures repertoire of functional groups and amino acids. Through genetic code expansion, it is now possible to augment nature's amino acid alphabet with additional functionalities. In particular, through integration of amino acids that are fused to small molecule catalysts, which have been extensively explored by organic chemists, it is now possible to create enzymes with a whole new reaction scope that is unprecedented in nature. In order to design better enzymes with non-natural functionalities from scratch, it is essential to gain fundamental understanding of how artificial amino acids must be placed and further complemented within enzyme active sites. This project aims to address this by experimentally improving designer enzymes with non-natural amino acids, leading to a fundamental understanding of the true potential of augmenting nature with additional building blocks. This will be achieved through directed evolution, which is a mimic of Darwinian evolution on a laboratory time scale. Directed evolution allows for the discovery of mutations that improve enzyme activity but are rationally not predictable. Through iterative rounds of mutagenesis and selection, the activity levels of enzymes can be significantly improved. This is a laborious process as many enzyme variants have to be individually analysed, to identify rare mutations with beneficial effects. To accelerate this process, an ultra-high throughput assay will be implemented that utilises picolitre-sized droplets as reaction vessels that contain the enzyme variant and its coding gene. Droplets can be manipulated at high speeds of several thousand droplets per second and most importantly they can be sorted according to enzyme activity. With this technology at hand, it is now possible for the first time to explore the evolutionary limits of these designer enzymes for different unrelated reactions, including ester hydrolysis and an unnatural carbon-carbon bond forming reaction. In depth characterisation of the best performing enzymes will highlight functional and structural features that are essential for supporting catalysis by nonnatural amino acids. Recapitulating these findings by computational enzyme design will challenge our gained molecular understanding and give rise to new generations of improved designer enzymes. Overall, this research project will open up new avenues in development of highly active enzymes for abiological reactions with implications in biotechnology, biocatalysis and synthetic biology.