NEW YORK – Enzyme engineering company Codexis has been developing an enzymatic RNA synthesis technology that it hopes will "alleviate concerns over the scalability, sustainability, and economics" of traditional phosphoramidite-based chemical RNA synthesis.
Dubbed Enzyme-Catalyzed Oligonucleotide (ECO) Synthesis, the method relies on engineered terminal deoxynucleotidyl transferase (TdT) to produce small interfering RNA (siRNA), which is double-stranded, de novo. Once the technology is ready, Redwood City, California-based Codexis plans to become both a service provider and benchtop instrument maker for enzymatic RNA synthesis to target different market segments.
"About two or three years ago, we started to hear questions from our partners around 'Can you use enzymatic techniques to do oligo synthesis, particularly RNA synthesis?'" said Codexis CEO Stephen Dilly. "What they were noticing was that phosphoramidite chemistry has been around for a long time, it's really good, it's really optimized, but it's really hard to scale."
Previously, Codexis engineered a version of TdT for enzymatic DNA synthesis, landing the company a partnership with San Diego-based synthetic biology firm Molecular Assemblies, which is commercializing its DNA manufacturing tech using Codexis' TdT.
However, Dilly noted that the DNA TdT is "mostly inactive on RNA," so the company had to "start from scratch" to develop a version of the enzyme specific for RNA synthesis about 18 months ago.
To engineer the enzyme, Dilly said Codexis introduced point mutations into the wild-type TdT, leveraging artificial intelligence and machine learning to inform its evolution.
Specifically, he said the goal was to produce a TdT that can grip the starter RNA oligo while adding the building blocks with minimal sequence bias. In addition, the enzyme had to accommodate RNA base modifications, particularly those at the 2' end of the ribose group that are commonly used in RNA therapeutics.
"It's all about complexity and promiscuity when you are constructing RNA, particularly siRNA," he pointed out.
Mechanistically, the process of ECO Synthesis is similar to TdT-based DNA synthesis, which involves cycles of extension and deblocking steps.
To ensure that TdT only incorporates one nucleotide at a time, Dilly said, nucleotide triphosphates (NTPs) are converted into nucleotide quadro-phosphates (NQPs) by adding a reversible phosphate blocking group to the 3' end. The blocking group is then removed by phosphatase-catalyzed hydrolysis during the deblocking step, allowing the TdT to incorporate another base during the subsequent cycle.
According to Dilly, so far, the RNA TdT has undergone 60 rounds of evolution, reaching a conversion efficiency in the 60 percent to 70 percent range. Overall, the company is expecting the enzyme to have "well above 90 percent" conversion efficiency, he added.
In addition, he said the company will need to enzymatically produce NQPs, which are "not yet widely available," and the starter oligos, which are currently produced chemically.
As Codexis continues to mature its ECO Synthesis technology, the company is making plans to apply it in different areas, Dilly noted.
For big pharmaceutical companies or contract development and manufacturing organizations (CDMOs), he said the company could license the ECO platform out, enabling these customers to produce siRNAs on their own.
For smaller manufacturers and pharmaceutical startups, Dilly envisions Codexis becoming a "very-rapid-turnaround service provider" that can produce good manufacturing practices (GMP)-grade siRNA at a smaller scale.
Eventually, Dilly said, Codexis plans to develop ECO Synthesis into a benchtop platform for researchers, adding that the company is already working with technology partners "to figure out how to scale."
Codexis is not the only company eyeing the enzymatic RNA synthesis market. EnPlusOne Biosciences, a Harvard University spinout cofounded by George Church, for instance, is trying to commercialize its so-called ezRNA technology. In addition, South San Francisco, California-based DNA Script, which already has an enzymatic DNA synthesis benchtop platform on the market, has also shown an interest in the RNA space.
Compared to competitors’ technologies, one strength of ECO Synthesis is its tethering approach, Dilly said, which anchors the TdT enzyme to the reaction column while the growing oligos and building blocks wash through during each cycle. "What we had to show is [that] we evolved the enzyme so that you can tether it without interfering with this activity," he explained.
He also believes the science around the phosphatase-based deblocking mechanism is "very elegant," and the company has "the IP dominant position" around using that blocking group.
Moreover, one more "secret weapon" for Codexis is its proprietary double-stranded RNA ligase, he said, which allows researchers to stitch together shorter RNA molecules to make a longer strand.
That enzyme is also applicable to customers who are currently using the chemical synthesis route, which has its strength in producing smaller RNA oligos, Dilly noted.
"There are some things that [chemical synthesis] is super good at — it's very flexible, very rapid for short sequences," he said. "I think the real question for us is not how we replace phosphoramidite chemistry, but how we fit in with it."
By the end of this year, Codexis plans to demonstrate the scalability of the ECO Synthesis technology by making "a reasonable quantity of an siRNA that is a real sequence, done under proper reaction conditions, and has a decent purity profile," Dilly said.
Early next year, the company aims to show that it can synthesize a full-length siRNA in one go. After that, Dilly said the company plans to sign the first commercial contracts for the technology with pharmaceutical partners in the fall, followed by further scaling of the technology in early 2025.
In addition, the firm plans a full commercial launch of its double-stranded RNA ligase in the middle of next year.