Machine learning-guided acyl-ACP reductase engineering for improved in vivo fatty alcohol production

Abstract

Fatty acyl reductases (FARs) catalyze the reduction of thioesters to alcohols and are key enzymes for the microbial production of fatty alcohols. Many existing metabolic engineering strategies utilize these reductases to produce fatty alcohols from intracellular acyl-CoA pools; however, acting on acyl-ACPs from fatty acid biosynthesis has a lower energetic cost and could enable more efficient production of fatty alcohols. Here we engineer FARs to preferentially act on acyl-ACP substrates and produce fatty alcohols directly from the fatty acid biosynthesis pathway. We implemented a machine learning-driven approach to iteratively search the protein fitness landscape for enzymes that produce high titers of fatty alcohols in vivo. After ten design-test-learn rounds, our approach converged on engineered enzymes that produce over twofold more fatty alcohols than the starting natural sequences. We further characterized the top identified sequence and found its improved alcohol production was a result of an enhanced catalytic rate on acyl-ACP substrates, rather than enzyme expression or KM effects. Finally, we analyzed the sequence-function data generated during the enzyme engineering to identify sequence and structure features that influence fatty alcohol production. We found an enzyme's net charge near the substrate-binding site was strongly correlated with in vivo activity on acyl-ACP substrates. These findings suggest future rational design strategies to engineer highly active enzymes for fatty alcohol production.

TransMEP: Transfer learning on large protein language models to predict mutation effects of proteins from a small known dataset

“…This is a well-known issue in optimization problems. 17 Similar to the work by Greenhalgh et al, 11 we use the upper confidence bound (UCB) criterion to achieve this trade-off. The UCB criterion is defined as…”

“…Machine learning has recently gained traction as an in silico tool to predict the expected effect of mutations based on known in vitro assessed mutants. [7][8][9][10][11][12] A machine learning approach to predicting mutational effects usually consists of two main components: First, features, i.e., a set of numerical values, are calculated for each mutant; second, a machine learning model is fitted to these features to predict the desired target value, such as the stability or activity of an enzyme. Several commonly used transformations are available for the features.…”

“…In the approach of Greenhalgh et al, a Gaussian process model was used that simultaneously predicts the mean and standard deviation of the target value. 11 This allows the application of a well-known optimization scheme, the so-called upper confidence bound optimization, an optimization method with sound theoretical support. 17 In upper confidence optimization applied to protein engineering, the user can control the trade-off between the exploration of the mutant space and the exploitation of interesting candidates.…”