Ben Chiarelli, Founder and CEO of Cellibre, a company revolutionizing agricultural manufacturing, discussed the company’s goal of shortening their complex design-build-test process to produce globally significant products at scale. Ben also shared how the partnership with Codex DNA empowered their team to rethink their enzyme engineering process. Hear from Ben about his experience improving throughput, both in the time to make the libraries and the speed of the engineering cycles.
What Cellibre does
We believe that biology is the most sophisticated and elegant manufacturing technology known to man, and we program biology to make products more sustainable, efficient, and reliable across various industries. Often people in the field of synthetic biology talk about sustainability. We don’t start there. Our key criteria for ensuring bio-based production are having large volumes, as biology is a scalable technology, so it’s all about large markets.
Secondly, we make sure we know what we’re getting ourselves into. Thirdly, we want to find supply chains that have major flaws. The goal is to solve issues and make products better, cheaper, and more convenient for the end consumers. We must ask ourselves if a biobased solution is economically viable or not. Can we compete with legacy supply chains in the marketplace to bring overall costs down? Our first application of this technology is in the cannabinoid space because it really does check all the boxes.
Cannabis industry focus
In large markets, most people think of cannabis as a recreational market like vape pens, beverages, etc. There’s also a large human therapeutic market as seen by evidence from GW Pharma and folks like that. There is also an animal health market for these molecules as well. We’re talking about the potential for hundreds of millions of kilograms in demand per annum across those three areas.
Additionally, their biochemistry is known. This plant has been sequenced. We understand the whole pathway, and we know how to turn those into valuable products known as cannabinoids.
Challenges with traditional agricultural methods
The supply chain for legacy agriculture is filled with problems. The plant itself makes over 400 chemicals simultaneously, so any product you get will be impure and inconsistent. The plant is a super sucker, so it takes everything it finds out of the soil and brings it into the plant, making it susceptible to contaminants like chemical residues or pesticides. There are also long unpredictable production cycles that go along with producing these molecules via traditional agriculture. Lastly, the prices for cannabinoids look a lot like medicines, not as commodities, like nutritional proteins or biofuels. The economics provide a floor and make the biobased solution more viable.
Cellibre changing the way cannabinoids are harvested
Traditional companies take an organism approach like the many great companies that have built tens of millions and in some cases hundreds of millions of dollars of infrastructure around engineering certain cell types. Whether that be saccharomyces cerevisiae, also known as bakers yeast, E. coli, or other bacterial systems. They typically do this by using fuels or specialty chemicals and taking that process and infrastructure to engineer cells to make various products. Essentially taking those square pegs and fitting them into round holes.
What we do differently here at Cellibre is rather than taking a cell approach, we take a product approach. If we want to make CBD we find out if there is a cell in nature that already makes the building blocks where the primary is to build the precursors that we need to make the product of interest. We then engineer those cells and allow nature to get us closer to the endpoint. This is the fundamental difference between us and other people trying to use biology for manufacturing and technology.
Codex DNA partnership improving the enzyme engineering process
Our partnership with Codex DNA starts not at the cell level, but down at the enzyme level. When we’re building these pathways and have different enzymes doing different chemical reactions, along those pathways, we want to both discover and engineer those enzymes. This can be a lengthy process. You design and build your constructs and once you have your constructs, you plate those cells and determine whether you have a proper build. After that, you do DNA amplification, which leads to enzyme production and analysis of those enzymes. Then you repeat the design-build-test process as you hear of from more engineering-focused focused disciplines.
When done traditionally, it’s a fairly long process. Because Codex DNA’s solutions provide specificity and accuracy in their platform, we were able to accelerate this process and eliminate plating steps because we knew exactly what we were getting from them. This saved us almost a full week on each one of the cycles, which is a big deal from a time, IP, and cost perspective. When you think about how most people do protein engineering today, they’re doing what’s called pooled libraries, and that means they take hundreds of millions of constructs and pull those into a single tube. In the substrates, they screen those enzymes and find the winners.
Ultimately, most people are only sequencing and characterizing the winners. That could be problematic because of traditional oversampling issues, full data sets are only available for what you sequence. We only fully understand what those winners do, we don’t understand why things didn’t work which can be just as important. It’s especially cost-prohibitive for smaller organizations to sequence everything and get those complete data sets. Additionally, to do that screening and sequencing requires high throughput.
Benefits of using Codex DNA’s library services
With Codex DNA’s technology, we’re able to individually build constructs to do everything de novo. We can build large libraries on a de novo basis, as well as do very targeted libraries of de novo constructs. Codex DNA’s technology can do this with a similar turnaround to pooled libraries. Additionally, the cost to do this is much lower than other technologies out there that we evaluated.
Once we have those libraries, we screen and find the winners. Interestingly, we can characterize both the winners and the losers because every single sequence is known as they were all built de novo. This means we get the same turnaround time and drastically lower costs of working with other providers. We also have the flexibility and fidelity to do both complex and simple designs. Because we have complete data sets, we can then flow back into our AI and machine learning algorithms, both for discovery and design. These clean data sets provide understanding not only into what works but what doesn’t work, which is crucial if you want to do in silico work that we as a field are so excited about.
Continue on for a brief case study on how this worked to get a sense of the different things that we were able to do with Codex DNA’s technology in this discovery and engineering program. First, one of the key enzymes in the pathway for making a cannabinoid is the CBG synthase. We discovered an enzyme that had great activity and when put in the substrate made a little bit of CBGA and a lot of other stuff. Although the activity was great, we would rather engineer specificity vs. activity.
Went through our first library builds doing traditional site-saturation mutagenesis (SSM) with Codex DNA’s technology we were able to flip that specificity the other way and even further to the point where it was almost 99% specific for the product of interest. In addition, we increased the activity. We then came to Codex DNA and said we don’t want to just do point mutations we want to change entire sections of this protein and build these types of more complex libraries.
We did that with the team at Codex DNA in very short order and resulted in almost a 5x increase in the activity of these enzymes while maintaining the specificity. Lastly, we did another round of engineering and we got it up to six times more active than what that original enzyme was. All of this was done from a kind of ideation discovery, all the way to these improvements, in less than three months.
Over the course of time, we’ve continued to work with Codex DNA and this enzyme is now ten thousand times more active than the original discovery. It was the flexibility and specificity that allowed us not only to do the simple high-throughput screening that we would normally do, but we were able to take more complex engineering approaches with Codex DNA that led to these very significant discoveries for Cellibre.
With that, I would like to thank the entire team at Codex DNA for their partnership in the continued innovation that they bring to us every day.