Please enjoy this interview with Zachary Sun from Tierra Biosciences! It originally came out in my weekly newsletter; you can subscribe here if you haven’t already.

Tierra is a phenomenally interesting company with whom I worked while I was at Social Capital and am lucky enough to get to stay in touch with. Tierra is blazing ground in a new field of biological manufacturing called Cell-Free Systems Biology, a really interesting new way to discover, test and engineer new molecules that do useful things for us in the world.

In a nutshell, biomanufacturing generally means engineering living systems, like cells, to mass-produce some sort of chemical product that’s useful. This is how we make crucial products like insulin. The good news about this strategy is that cells are already very good at making these things, so a large part of biomanufacturing is figuring out how to get these cells to do what they naturally “want to do” in controlled conditions. The hard part is that living systems are incredibly complex. It can be very hard to understand what’s going on when so much of what happens inside a cell is outside our realm of understanding.

Cell-free systems biology, where Tierra is a pioneer, is a potential solution for that problem: figuring out how to strip away most of the parts of the cell, while preserving a few essential features, in order to simultaneously increase our control and knowledge over what’s going on in our engineered system while also increasing the sphere of possibility for what we can make. The idea is to engineer a system that takes advantage of the basic functional capacity of a cell, without necessarily using the cell.

If you like, you can think of cell-free systems and other stripped down biological platforms as a little bit analogous to breadboards we can build hardware setups on, or maybe even the very first microprocessors like the Intel 4004. The earliest “computer-on-a-chip” processors were severely underpowered compared to their mainframe and minicomputer competitions, who already handled existing workloads for enterprise customers not unlike the way that drug manufacturers already use engineered cells to do jobs like synthesizing insulin today. But their simplicity was also powerful: writing software on top of generally programmable hardware changed the world of computing, even if it seemed simple at first. Could stripped down cell-free systems open a similar door for the world of biology? Maybe; we’re sure going to try.

Tierra is quite early-stage, but they’re a company that gets me particularly excited and I love thinking about the sheer amount of potential that Tierra and others in their field are capable of unlocking. Zach has graciously offered to chat with us this week about it.

Why does the world need Tierra Biosciences? What are you building, and what are you trying to achieve with it?

At Tierra, we care a ton about finding new molecules from Nature. We rely on natural products every day – think antibiotics, chemicals, herbicides, or even the enzymes in laundry detergent. We make these now using biomanufaturing. However, we’re also finding it harder and harder to find new molecules, even though we know Nature has so much biological and chemical diversity out there to mine – some 1 trillion microbes exist by some estimate. Problems like antibiotic resistance are the result of our inability to find new molecules.

Nature has the instructions to make its molecules in DNA, and at Tierra we’ve built the world’s fastest translation engine to take that instruction set and re-create those molecules. We’re doing this by removing the cell out of the equation, which is called a cell-free system. This is pretty important, since many of the instructions come from microbes we don’t even know, let alone know how to grow – from all that next-generation sequencing data we’ve collected that blueprint the 1 trillion organisms out there. We accept that biology is complex; we resign ourselves to the fact that we may not know what (else) we need to make this molecule other than the DNA. To get around that, we make every cycle through our system as fast and efficient as possible; we collect data on every run to learn from both successes and failure; and we utilize different cell-free chemistries to represent different biological complexity, to experimentally determine what works (and do learning to make the next cycle better). This is at its core hard-tech – we make the molecules, not just simulate them – so we can actually test things, not just speculate.

Tierra is part of a new wave of businesses that I think of as “21st century manufacturing companies”, where what really sets you apart among other tech companies is that you want to make real, physical product. They’re very small products, but they’re still physical products. For the uninitiated, what are some of the steps and processes that go into biomanufacturing, however you’d like to define that, and what’s different than it was 5-10 years ago?

The concept of bio-manufacturing, let’s define that here as using a microbe (or components thereof) to make a product, is not necessarily new; we’ve been converting rice into alcohol through fermentation for 5,000 years, and converting sugar into insulin for 40 or so.

At its core, you need an input (or food source). Typically, this is sugar, but it can also be waste products (think flue gas or agricultural waste), or even light if you’re using algae. The microbe (or components thereof) take in this food source and convert it to a product through a genetic instruction set that is either native (think yeast to alcohol) or programmed in by humans (think using bacteria to make insulin). It take a good amount of time to program in the genetic instructions, and then to make yields high enough to make it worthwhile to do for most products.

It’s only in the past decade we’ve really had the ability to “turbocharge” the process per-se – to be able to sequence DNA to be more sophisticated of what to engineer, to be able to synthesize DNA as those building blocks, and to be able to implement automation and genome engineering tools (think CRISPR). We’ve also been humbled a bit; in the late 2000’s, as a field we thought we could make anything using this bio-manufacturing process, including oil, only to realize the engineering cycles are quite long and biology is still fairly unpredictable. I’d say the past 10 years has seen a balance of more abilities with an acknowledgment that costs (in time and money) for doing bio-manufacturing can be quite high. We’re seeing now many more successes by companies selectively choosing items to bio-manufacture with high margins, like perfumes and fragrances.

“Cell-free biology”, if you were to sum it up, is really all about taking the highly complex system that is the cell and stripping it down to something more manageable that still has some basic function intact. How do you think about the “job” that you want cells to do?  

By using cell-free biology, we’re trying to balance out accepting the parts of a cell we may not want to, or be capable of understanding – all the complexity and regulation that goes on in the background – while still conducting engineering. The goal here is to be able to answer your hypothetical questions (eg. Does this pathway make a molecule? Does this piece in a genetic circuit work?) in the most similar context of a cell that doesn’t involve actually engineering a cell, since engineering a cell is quite painful.

For our applications, cell-free systems do the job of the cell we’re most interested in – translating an instruction set, encoded in DNA, into a product, either proteins or small molecules that proteins create, and providing the “supportive environment” (the things in the cytoplasm composed of stuff we may not necessary understand) to make this all happen. Every cell-free system relies on the classical central dogma, transcription and translation, to do the first part. The “supportive environment” part, and composition of the cell-free system, will range depending on what microbe it’s made from and how we process the microbe into the cell-free system.

In this sense, we really just use cell-free systems as a data collection method that produces our desired product. We get the cell-free system to do the translation step through a combination of rapid testing cycles and learning from successful and failed cycles. That data can then be ported to cells, which are great at self-replicating in a way cell-free systems can’t; or can remain in cell-free if the item produced is toxic or otherwise hard to engineer for cellular production.

So, when I took molecular and cell biology in college ten years ago, we were taught a way of thinking about biology that is now quite out of date. We were taught this pretty rigid viewpoint of Central Dogma of Cell Bio (DNA —> RNA —> Protein), where RNA was basically just this temporary placeholder for information and that’s it. We were taught a pretty traditional concept of the cell. So my question is: what do you think has been, or ought to be, the biggest shift in our basic conceptual understanding in how we think about and how we teach molecular & cell bio?

It’s a good question. My molecular biology class was a little bit longer (15 years now?) but sounds like we learned a lot of the same “simplistic” understanding! I’m not sure how it’s taught today; I fear that it may be more of the same, but I think people are starting to appreciate more that the cell is a lot more complex that a model can predict; and even things that are simple, such as the central dogma, are not as rigid as they seem as we figure out the multiple uses of RNA.

Perhaps we’re seeing a humbling of our ability to really fully “understand” bio, and we’re starting to really see how complex biology really can be. When you build things by evolution (like biology does) rather than rational design, this starts to make sense. A humbling moment for me was seeing all the excitement of sequencing the human genome (and a ton more genomes), only to find that regulation is way more complex than we thought before; that “junk DNA” is not junk DNA at all, but rather things we may not understand.

I’m hoping that appreciating this complexity a bit more is a change, though again, not having taken a molecular bio class recently, it’s hard for me to say.

I worked in a lab when I was a graduate student, but I’d be pretty lost in many modern labs today, simply because the tools we have available have accelerated so much: they’re so much more powerful, and have automated or facilitated so many of the routine tasks that used to take dozens of hours of somebody’s time. What are a key handful of technological accelerants that have really changed bench work and make what Tierra does possible?

This is an interesting point; actually, despite all of the tools out there I’m still surprised by how much of a “drudge” factor there is ongoing in labs. The tools that are developed are developing at such fast pace that even for someone at the cutting edge it’s really hard to be up-to-date at all times in all fields, and since wet lab science is so interdisciplinary this challenge becomes even harder.

A good example is in cloning DNA; only in the past 8 years have we seen game-changing new techniques arise (eg. Golden Gate, Gibson Assembly), as well as the ability to print DNA (eg. Twist) and manipulate small quantities of reagents extremely precisely (eg. Labcyte Echo). These techniques, services, and instruments have really changed the way we run our whole platform front to back. While these techniques may be known to those steeped in the field in synbio labs, they also impact labs in other biology fields that will inevitably use molecular biology techniques, but they’re all still relatively niche.

There’s also been a push to automate cloning, by platforms such as Transcriptic and UW Biofab. However, cloning itself can be so complicated that inevitably humans have to get re-involved to get things past the finish line.

One really interesting thing happening in synbio right now is the creation of a “stack” per-se, an infrastructure layer that help people implement cutting-edge tools in a streamlined manner. I’m thinking integration between DNA Cad tools (eg. Genome Compiler, now in Twist; and API hook-ins) to delivery of DNA in Echo plates that allows seamless usage on automated platforms. Hopefully these types of things will lead to standard workflows that really save people time in the field.

There’s been a general sense that Bio could be the next major frontier for early stage investing and building, like where software or the early internet was a few decades ago. I think that’s accurate, but right now most of the value we’re building now is still quite forward-looking, speculative, and requires some leaps of imagination to get right. A lot of these new bio companies aren’t doing drug development, where it’s like, “if we successfully invent a new Alzheimer’s treatment, we’ll make a billion dollars, and if we don’t, we make zero dollars.” There are questions around product market fit; there are researchers and companies pondering, “we’ve built something interesting, but we aren’t sure what it’s good for yet.” How would you describe the challenges and opportunities facing an early-stage bio company today, and how are they similar and different to software startups, or old-school biotech companies?

This is probably the question that keeps me most up at night. Finding product market fit is one of the biggest challenges of startups in this field, especially as these new technologies have the ability to be quite disruptive to current markets. For us at least, finding initial market traction for utilizing our tech has been pretty critical to validating a “use case.” Balancing that traction with a fundamentally disruptive thesis – that molecules are not found correctly, that new technologies will change the way we find new matter – will be quite important moving forward.

The other point here worth mentioning is that biology is fundamentally expensive to do, unlike software now (but probably like software and hardware in the heyday of the internet). Government support for pushing these technologies to the point of commercialization has been really critical. For us, it’s been pretty critical to give us time to hone our platform while identifying product-market fit, as in a bio-space, which is inherently hard tech, you really have to do both to be successful.

The path to Product Market Fit for synthetic bio companies like Tierra is hard, but we’re all here because we believe the potential impact can be enormous. Building these early platform technologies, like early computing or the early internet, hopefully pay off big time when we reach a tipping point and see an explosion of new applications built on the fruit of many years of labour. If you peer into the future, what do you think will be the biggest applications of synthetic and cell-free biology that people may not often think about, and what do you think will be Tierra’s biggest impact on the world should you succeed?

Good question. I’m pretty bullish on cell-free biology, and wrote an article on what I see the field can do. When I think of Tierra in specific, I think our contribution will be two-fold. In the long term, we’re aiming to help bring new molecules to market, and we’ve already seen cases where we’ve been able to accelerate the discovery process with our commercial partners. Again, we’ve seen what natural products can do in the past, like the antibiotics revolution, so one can imagine the impact that would have on society in the future. In the short term though, at Tierra we’ve really been pushing the envelope on cell-free engineering. We’re already getting those in the synthetic biology field itself to think outside-the-cell per-se, and accelerate their engineering cycles. By speeding up the design-build-test cycles, this should help accelerate the R&D happening in the whole synthetic biology field.

Big thanks to Zach for the thoughtful answers, and if you’re a biology graduate who’s looking for a really awesome way to make the jump from the lab to the startup world, they’re currently hiring for a Research Associate position in the Bay Area. If you or someone you know is looking to make that career evolution, you should jump at this chance: this would’ve been a perfect job for me out of grad school and I’m sure there are some of you for whom that’s also the case. Take a look, and forward along to friends who might be interested.