One of the grand challenges in synthetic biology is building enzymes that do not exist in nature, and making them work inside living cells. Artificial enzymes, constructed by embedding synthetic catalytic cofactors into protein scaffolds, have been developed in test tubes for years. But the crowded, chemically complex environment of a living cell has proven a formidable barrier. Diffusion is restricted, reactive side-products abound, and conditions are far from the carefully controlled buffers of an in vitro reaction.
A team led by researchers at Jiangnan University and Xiamen University in China has now cleared that barrier. In a study published July 2 in Nature Communications, they report the first efficient strategy for assembling artificial enzymes directly inside living cells, where the enzymes catalyze asymmetric carbon-carbon bond-forming reactions with excellent stereochemical control.
“The complex cellular environment poses numerous challenges for artificial enzyme design,” said corresponding author Zhi Zhou of Jiangnan University’s School of Life Sciences and Health Engineering. “Our approach mimics how natural enzymes are matured: the cells express a protein scaffold, then we incubate them with a synthetic cofactor that anchors itself in place.”
How it works
The strategy is elegant in its simplicity. The researchers engineered bacterial cells to express a target protein containing a strategically placed cysteine amino acid. They then added a synthetic catalytic cofactor, a small molecule designed to perform a specific chemical transformation, to the cell culture. The cofactor diffused across the cell membrane and formed a site-specific disulfide bond with the cysteine residue, locking the cofactor into the protein scaffold.
The result was a functional artificial enzyme assembled entirely within the cytoplasm, performing an asymmetric Mannich reaction, a fundamental carbon-carbon bond-forming transformation that is a cornerstone of organic synthesis, used to build complex molecules with defined three-dimensional structure.
The reaction achieved excellent enantioselectivity, meaning the enzyme consistently produced one mirror-image version of the product over the other. For pharmaceutical chemistry, where the two mirror-image forms of a molecule can have entirely different biological effects, enantioselectivity is the difference between a drug and a toxin.
Crystallographic analysis and computational studies, including density functional theory and quantum mechanics/molecular mechanics simulations, revealed the structural basis for the stereoselectivity. The protein scaffold positioned the cofactor in a precise orientation, creating a chiral environment that favored one reaction pathway over its mirror-image alternative.
A generalizable platform
The key claim of the study is that the approach is generalizable. Because the cofactor anchors through a simple disulfide bond, a well-understood and widely used biochemical linkage, the same strategy could be applied to different protein scaffolds and different synthetic cofactors. The researchers tested their system with a specific Mannich reaction catalyst, but the underlying principle, express a scaffold, add a cofactor, let it self-assemble, is modular.
“The method does not require any special reagents or genetic engineering beyond placing a cysteine at the right position in the scaffold,” said corresponding author Binju Wang of Xiamen University. “That means it could be adapted to many different reactions simply by swapping the cofactor or the protein.”
This modularity opens the door to building libraries of artificial enzymes inside cells, each catalyzing a different reaction, without the need to purify, reconstitute, or deliver pre-assembled enzyme complexes. For applications in metabolic engineering and biosynthesis, where multiple enzymatic steps must occur sequentially within a single cell, the ability to install artificial enzymes on demand is a significant advance.
Why asymmetric Mannich reactions matter
The Mannich reaction is one of the most important carbon-carbon bond-forming reactions in organic chemistry. It produces beta-amino carbonyl compounds, which are building blocks for a vast range of natural products and pharmaceuticals. The asymmetric version, producing only one enantiomer, is especially valuable because biological systems recognize molecules by their three-dimensional shape.
An artificial enzyme that performs this reaction inside a living cell, with the stereochemical precision of a natural enzyme, provides a tool for producing chiral intermediates directly in a biosynthetic pathway. Instead of synthesizing a chiral molecule in a chemical plant and then feeding it to engineered microbes, the microbes could produce the chiral center themselves.
Caveats and next steps
The current system has limitations. The disulfide bond anchoring the cofactor is susceptible to reduction in the cytoplasm, which could limit the enzyme’s lifetime and require ongoing cofactor replenishment. The reaction scope, while promising, has been demonstrated for a specific substrate class. The researchers note that extending the approach to other reaction types, oxidations, reductions, cyclizations, will require cofactors with different catalytic properties.
The demonstration that artificial enzymes can be assembled inside cells at all, however, represents a step change in the field of in-cellulo biocatalysis. It moves the field from “can we build an artificial enzyme?” to “where can we deploy it?”
Source: Zhu Z, Hu Q, Wu Y, Wang B, Zhou Z. Intracellular assembly of artificial enzymes for cytoplasmic enantioselective Mannich reactions. Nature Communications (2026). DOI: 10.1038/s41467-026-75059-9
This article was translated and adapted by the Babel translation pipeline.