Enzymes are powerful catalysts that are used in a range of industries, from biomedicine to manufacturing. In laboratory settings, naturally occurring enzymes have unlocked important advances such as PCR and CRISPR, and offer great potential for green chemistry. Pushing these advances further requires the creation of enzymes that are not found in nature, ones that can perform non-natural chemistries to deliver a range of desirable functions and products. However, their design is still an arduous process, limited by computational power and labour costs. The EU-funded ProFF project sought to address these challenges with a new system for the directed evolution of enzymes. “There are a lot of applications for which you need enzymes that you cannot find in nature, enzymes that are more efficient, more resistant to temperature, and so on,” explains project coordinator Yannick Rondelez, researcher at the ESPCI and CNRS. Rondelez, whose background lies in molecular programming, a form of synthetic biology, took the approach of designing biological machines that can operate without a living host, encoded by small short stretches of synthetic DNA called oligonucleotides. “We don’t create these for biochemical purposes, but for computational purposes,” says Rondelez. The systems exploit the fact that the evolutionary process can act as an algorithm, selecting for specific outcomes over repeated iterations – in this case, improvements to enzymes.
The typical approach to protein engineering involves creating multiple variants and testing their activity one by one. The best are selected, and the process repeated. This makes it a slow and labour-intensive process. “The larger the library of variants you can manipulate, the higher the chance there is to find a good one,” adds Rondelez. “But it’s difficult to do this traditionally at high numbers. We’re trying to remove this bottleneck, and manipulate very large numbers of enzymes.” To do this, Rondelez and his colleagues sought to create a molecular program that connects the phenotypic activity of an enzyme with genetic replication. “We make an emulsion with a billion gene compartments, each contains one variant and the molecular program,” he remarks. “If the enzyme has correct activity, it triggers the molecular program and the replication of its own gene.” The approach exploits various phenotype-genotype linkage strategies, as the performance of an enzyme must be tested while simultaneously keeping track of the genetic sequence that produces it.
“In our project we worked a lot on enzymes where the substrate is DNA itself: ligases and so on,” explains Rondelez. “Because our molecular program is made out of DNA, it’s easy to connect enzymatic activity to a DNA-based system. But it’s important to note that we showed the approach was also possible with enzymes acting on non-DNA substrates.” The project was supported by the European Research Council. “This helped me to equip a new lab and hire a new team,” says Rondelez, who moved from Japan to France to head the project. “It was a very risky move to integrate many techniques that we’re not specialists of, and that was something that can be done only with large, long-term financial support.” A number of papers have been published or submitted for publication, and one aspect of the research has led to the development of a spin-out diagnostics company. Rondelez says he plans to make the technique his team has pioneered more accessible and faster, allowing enzymes to be optimised in a matter of days.
ProFF, enzyme, directed, evolution, oligonucleotides, DNA, diagnostics, linkage, genotype, phenotype, replication, activity