What they're after: Every plant, animal, and microbe functions because of enzymes. They catalyze the digestion of food, prevent blood from clotting, and help cells communicate with one another. Wine, bread, cheese, medicines, contact lens solution, laundry detergent, and a long list of other things we enjoy, dislike, and otherwise rely on work because of enzymes. They are, in other words, vital, useful, and lucrative (industrial enzymes are a nearly $5 billion business).
Tens of thousands of enzymes have been isolated and described but the challenge in finding new reactions is two-fold:
- First, enzymes have to be identified. The relatively cheap cost of genetic sequencing is allowing companies to build vast libraries of genetic sequences for enzymes found in microorganisms that work in extreme environments — high or low temperatures, acidic, or basic solutions — similar to those in industrial processes.
- But, enzymes have been honed through evolution to perform specific functions in a particular organism, and it can be difficult to use them beyond nature's intent.
How it's done: For nearly 30 years, engineers have used directed evolution, a technique pioneered by Caltech's Frances Arnold, to modify enzymes for industrial use.
- It involves introducing random mutations into a gene to produce new versions of the protein it encodes, which are then screened for desired properties.
- It's been used to make new chemistries altogether: Arnold's team created an enzyme that can form carbon-silicon bonds that are rare in nature.
- The powerful approach has improved and optimized detergents, biofuels, the synthesis of drugs and a host of other products and processes, but it can't solve every problem and it can be slow.
This is where startups see an opening.
Some key players: For their first foray into enzymes, Ginkgo Bioworks is partnering with food ingredient company Kerry and Swissaustral, which develops extremophilic enzymes for scientific research, personal care, and chemicals. The team is engineering strains of microorganisms to produce industrial enzymes at scale. They plan to later screen proprietary collections of enzymes (like temperature-resilient catalases that protect cells from oxidative damage by breaking down hydrogen peroxide) and attempt to engineer them for food and textile industry applications.
Ginkgo's automated "foundry" speeds up the time to develop the strains of microorganisms for producing an enzyme. San Francisco-based Zymergen is also using machine learning and robotics to manipulate microbes en masse into making materials.
Seattle-based Arzeda takes a different tack: They start with what they need an enzyme to do — cut this, add that — and work backwards, using computer algorithms to design enzymes and pathways with the function they want. They then use Arnold's method of directed evolution to hone those proteins. For INVISTA, Arzeda created a way to synthesize nylon precursors using sugar as the starting point.
"We design proteins that don't exist in nature and expand the repertoire of natural enzymes in order to make new reactions for industrial applications. The possibilities are truly limitless to expand on what nature has evolved and create new function at the protein level," says Arzeda CEO Alex Zanghellini.
DuPont, Novozyme, Cargill, BASF and other heavy-hitters in the chemicals industry that manufacture enzymes — and use directed evolution themselves — are potential customers for the startups. But so far, industries have been conservative in adopting synthetic biology, in part due to regulatory unknowns and concerns about public acceptance.
"I don't think we'll be limited by the complexity of biology. It will be more limited by the pace at which we can discover commercially and socially relevant jobs and problems we need done. That is the rate-limiting step," says Brian Brazeau of Ginkgo Bioworks.
