Sustainability, profitability, and a radical shift in manufacturing
A lasting environmental shift, says CSO Zach Serber, is compatible with our desire for better products and a more fulfilling life.
“Humankind is in trouble,” says Zach Serber, Zymergen co-founder and Chief Science Officer, in a break from his usual optimism.
Everyone wants to live a better, easier, more fulfilling life, he says. But that often means consuming more from a planet whose carrying capacity is stretched to the limit. “The only way to improve people’s quality of life and reduce their impact on the environment is through biology,” he says. “Biofacturing is the answer.”
Where business and sustainability thrive
Zach and his co-founders launched Zymergen in 2013 to be an environmentally sustainable company. Their goal is and always has been to make better products in a better way. But they also knew that they needed to do more.
“To create a lasting environmental movement and a thriving bioeconomy, one has to create an economic model behind it.”
“To create a lasting environmental movement and a thriving bioeconomy, one has to create an economic model behind it ,” he says. “Biofacturing harnesses people’s natural desires for higher quality of life to make products in a better way, so that sustainability need not be a fight against human nature or capitalism.”
The alternatives to a bio-based future — lowering our quality of life or dramatically reducing the population — are much less appealing. So what’s been holding us back?
Turning the crank faster on the bioeconomy
For the bioeconomy to flourish, Zach says the process of creating product-making microbes has to happen routinely, quickly, and inexpensively. If there’s too much technical uncertainty, it will seem like a risky investment and hold the whole sector back.
“Today’s biomanufacturing has largely taken a mechanistic approach to making products,” Zach says. By that, he means that the traditional approach is to deeply study the metabolism and physiology of microbes, then try to put different genes together for a specific purpose. The idea is that if you understand how the genes and elements of the cell all work together, you can design a system to suit your needs.
The problem, Zach says, is this: what we don’t know about biology, genetics, and physiology vastly exceeds what we do know. “With a mechanistic approach, you can intervene productively sometimes. But there are clearly opportunities for making improvements outside the realm of what we humans understand.”
A glance around the natural world provides plenty of evidence for that. Without the benefit of science or ingenuity, nature has evolved exquisite solutions to countless problems. But whereas nature might take thousands of years to solve a problem, humans seek to design similar solutions in just months, weeks, or — as in the case of a COVID vaccine — mere days.
A radically empirical approach
As Zach started thinking about how to solve this problem, he took a lot of inspiration from people like Frances Arnold and Richard Fox, who worked on the directed evolution of enzymes and proteins. They found that, even without knowing the exact atomic structure of a protein, you could still improve its function by systematically changing each amino acid in the protein for one of the other 19 available. By doing this, they explored the design space methodically to find beneficial changes and then thoughtfully combine them to make better enzymes and proteins.
“That same approach ought to apply to an entire genome,” Zach says. “But at that scale, the experiments are too big. There are just too many amino acids and proteins.”
So Zach and team developed a way to improve the microbe at a slightly higher level using a series of proprietary libraries. One of the first they invented was a library type they call “promoter swapping” whereby they replace the regulatory region upstream of a gene with promoter regions of varying strengths.
“This allows us to turn the volume up or down on various genes. When we do this, we see that over- or under-expressing certain genes or pathways may have an unintended positive or negative impact on performance. Once these hotspots are implicated for one reason or another, we then can apply additional libraries to hone in on them to make improvements even greater.”
Now, they could scan an entire genome for changes that improve performance, whether or not they yet understood how those changes worked. What they found was surprising.
“Most of the improvements are not explainable,” he says. “There are impactful changes in genes for an unrelated function, or sometimes we find new genes of unknown function.” In other words, the mechanistic tools of traditional biomanufacturing might be missing most of nature’s best stuff. “It’s also a humbling reminder of just how ignorant we still are about how biology works,” he says.
Unsurprisingly, recent work has confirmed that many gene edits with positive effect on a desired phenotype are unexplainable, underlining the importance of a radically empirical approach.
Using this radically empirical approach, Zach says they can take huge amounts of data, apply machine learning, and convert it into recommendations about the best designs to create. “That turns the traditionally slow, expensive, and risky process of making things with biology into something that has the potential to be routine, inexpensive, and reliable,” Zach says.
Turning academic pursuits into real-world change
In an alternate reality, Zach became a professor.
“I wanted to devote my life to pushing the frontiers of science and educating the next generation,” he says. “It’s what my father had done, and it seemed like a good, fulfilling life.” In his studies, he became keenly interested in the ability for biology to make fuels. “It seemed like a great way to wean the world of petroleum,” he says. So in 2007 he left academia to pursue biofuels.
“In retrospect,” says Zach, “making fuel from biology is an odd idea. Biology has this exquisite control over chemistry, but the idea of making exquisite chemicals only to burn them in internal combustion engines is a mismatch of the power of the technology.” And ultimately, he says, it proved to be economically incompatible with the petroleum industry: fossil fuels has a 100-year head start, with trillions of dollars already invested in refineries and infrastructure.
“They’re gonna fight you tooth and nail to keep you out. And so, I think it’s fair to say that [the first generation of biofuels startups] failed. All those biofuels startups of the aughts have either pivoted or gone away.”
That’s when the inspiration for Zymergen struck. It was something like this: the technology stack is immature but pretty cool, but making it work takes the right business model and a better technology platform.
Beyond robots and drop-ins
Zach and his co-founders had two key insights they believe distinguish them from those who came before in the quest to apply biology to manufacturing. The first was around automation.
“Like many others,” Zach says, “we were bought into the use of robotics to increase productivity, improve data quality, create safer working conditions, and so on — that was nothing new. Our insight was to couple robotics to data collection and machine learning so as to automate the generation of designs.”
So whereas labs had begun automating physical labor, Zach and his team also aspired to automate hypothesis generation — the design side of the process. This, he says, was new territory for the industry.
The other key insight was around drop-in replacements . Specifically, Zach and his co-founders believed that focusing on drops-in — as others had done before us — was a grave error.
“This was basically the reason that biofuels failed,” he says. “When you go after the same products others have already made from different sources, you’re competing against established markets. The current suppliers have a lot of influence, and the molecules are made cheaply and easily from petroleum. The likelihood that you can compete with them on cost is very low.”
Zach says the Zymergen insight was to simply embrace the difference that biology offers by looking for places where biology can make better products that people want and will pay for. “That is how we believe the bioeconomy will be built,” he says, “not by creating green alternatives to existing products, but by creating better products that take advantage of the virtues of biology.”
“That turns the traditionally slow, expensive, and risky process of making things with biology into something that has the potential to be routine, inexpensive, and reliable.”
On being a CSO
To make biofacturing work requires the quiet orchestration of several deep technical disciplines, says Zach.
“One of the great joys of my job is acting as an orchestra conductor, trying to get these different groups to work productively to make beautiful music together. Our artists include biologists, microbiologists, fermentation engineers, data scientists, machine learning experts, material scientists, chemists, automation engineers, and many, many others.”