A major challenge to achieving industry-scale biomanufacturing of therapeutic alkaloids is the slow process of biocatalyst engineering. Amaryllidaceae alkaloids, such as the Alzheimer’s medication galantamine, are complex plant secondary metabolites with recognized therapeutic value. Due to their difficult synthesis they are regularly sourced by extraction and purification from the low-yielding daffodil Narcissus pseudonarcissus. Here, we propose an efficient biosensor-machine learning technology stack for biocatalyst development, which we apply to engineer an Amaryllidaceae enzyme in Escherichia coli. Directed evolution is used to develop a highly sensitive (EC₅₀ = 20 μM) and specific biosensor for the key Amaryllidaceae alkaloid branchpoint 4’-O-methylnorbelladine. A structure-based residual neural network (MutComputeX) is subsequently developed and used to generate activity-enriched variants of a plant methyltransferase, which are rapidly screened with the biosensor. Functional enzyme variants are identified that yield a 60% improvement in product titer, 2-fold higher catalytic activity, and 3-fold lower off-product regioisomer formation. A solved crystal structure elucidates the mechanism behind key beneficial mutations.

Sorting large libraries of cells for improved small molecule secretion is throughput limited. Here, we combine producer/secretor cell libraries with whole-cell biosensors using a microfluidic-based screening workflow. This approach enables a mix-and-match capability using off-the-shelf biosensors through either coencapsulation or pico-injection. We demonstrate the cell type and library agnostic nature of this workflow by utilizing single-guide RNA, transposon, and ethyl-methyl sulfonate mutagenesis libraries across three distinct microbes (Escherichia coli, Saccharomyces cerevisiae, and Yarrowia lipolytica), biosensors from two organisms (E. coli and S. cerevisiae), and three products (triacetic acid lactone, naringenin, and L-DOPA) to identify targets improving production/secretion. Significance The design–build–test cycle of metabolic engineering is swiftly becoming test limited owing to advances in DNA synthesis and library design. At the same time, improving extracellular production requires high-throughput studies to detect extracellular (not intracellular) content. Traditional approaches that utilize chromatography-based analysis or plate-based fluorescence assays are relatively throughput limited and require a high degree of liquid handling. These limitations are avoided here by creating a generalizable approach for small molecule production screening. Namely, this paradigm combines “off-the-shelf” biosensors with large library screening from producing cells in a cell-type agnostic manner.

Halogen-containing molecules are ubiquitous in modern society and present unique chemical possibilities. As a whole, de novo fermentation and synthetic pathway construction for these molecules remain relatively underexplored and could unlock molecules with exciting new applications in industries ranging from textiles to agrochemicals to pharmaceuticals. Here, we report a mix-and-match co-culture platform to de novo generate a large array of halogenated tryptophan derivatives in Escherichia coli from glucose. First, we engineer E. coli to produce between 300 and 700 mg/L of six different halogenated tryptophan precursors. Second, we harness the native promiscuity of multiple downstream enzymes to access unexplored regions of metabolism. Finally, through modular co-culture fermentations, we demonstrate a plug-and-play bioproduction platform, culminating in the generation of 26 distinct halogenated molecules produced de novo including precursors to prodrugs 4-chloro- and 4-bromo-kynurenine and new-to-nature halogenated beta carbolines.