Towards novel enzyme functions: Insights from engineering promiscuous esterases and hydroxynitrile lyases in the alpha beta hydrolase fold superfamily

Abstract

Enzymes are biological catalysts that accelerate the chemical reactions that make life possible. Many enzymes are extraordinarily specific, catalyzing their native reaction with only one or two substrates. Perhaps paradoxically, most enzymes are also capable of catalyzing at least one side reaction, be it with an alternative substrate or, less commonly, a different type of reaction. These promiscuous activities are usually so slow and inefficient that they have no effect on the organism’s physiology. Nonetheless, catalytic promiscuity is the primary driver of new enzyme function because it is much easier to repurpose an existing, albeit slow, catalytic activity than to develop a novel function do novo. Leveraging this catalytic promiscuity, there has been great interest in using enzymes for applications as diverse as drug development, industrial chemical synthesis, and even plastics recycling. Oftentimes, enzymes need to be engineered for altered activity, a process that involves making mutations that lead to changes in the desired trait. Enzyme engineering requires careful consideration of several factors to be successful; protein stability, interactions between mutations, and the choice of engineering strategies can all affect the outcome. Here we describe our efforts to engineer two promiscuous enzymes for novel activities. First, we used rational design to increase the promiscuous ester hydrolysis activity of a hydroxynitrile lyase (HNL). By analyzing the correlated movements of residues within the enzyme in combination with molecular dynamics simulations, we predicted eight mutations that increased the catalytic efficiency of the enzyme roughly 1,400-fold. Yet the improvements in turnover (kcat) were more modest, showing a 10-fold increase for the same variant. Further, we improved the turnover (kcat) of the enzyme by designing a series of variants with an increasing number of mutations. Our best variant exceeded the kcat of the benchmark esterase by more than two-fold, a 1,240-fold improvement over the wild-type enzyme. Second, we engineered an esterase to improve its promiscuous activity against polyethylene terephthalate (PET), one of the most common plastics in use today. We created two small libraries of variants containing all possible combinations of mutations and tested their thermostability and activity against PET at different temperatures. Surprisingly, we found that activity was temperature dependent but that thermostability did not correlate with activity. Epistatic interactions between mutations were pervasive, an indication of the complex and often unpredictable nature of enzymes. Our findings in both projects have provided insight into the challenges associated with engineering enzymes for novel activities, as well as illuminated strategies for the development of new activities in the future.