"Holy T7 Polymerase, Batman!": R&D Considerations For More Awe-Inspiring xRNA

By Anna Rose Welch, Editorial & Community Director, Advancing RNA

We may never be 100% certain if Michael Koeris, associate professor of bioprocessing for the Keck Graduate Institute, is Batman disguised as an mRNA expert. However, as I explained in the first- of this two-part article, I started to think more deeply about the parallels between medicines and superheroes following a conversation with Koeris about where the mRNA space could stand to become “braver.”

Though mRNA has yet to become the subject of a comic book, she is no stranger to donning a cape or two, thanks to her triumphant (re)emergence from the shadows during the pandemic. It goes without saying we have some pretty ambitious therapeutic goals for linear mRNA and next-gen saRNA and circRNA. However, as my conversation with Koeris explored, we still have a lot of scientific and technical work/innovation ahead of us to make our RNA products “stronger” (in more ways than one).

In addition to sharing his perspectives on where we could be more innovative in delivering our mRNA/RNA in part one, in this follow-on Q&A, Koeris shares a few brief thoughts on where our mRNA R&D is ripe for innovation to transform our mRNA/RNA into more heroic products for our patients.

ARW: In the first part of our discussion, you emphasized the importance of not only improving LNP production, but also argued that it is in the industry’s best interests to start exploring alternative delivery methods. If we shift to drug substance production/IVT, where do you see room for growth in the current IVT process?

Koeris: Basically, all IVT reactions require an RNA polymerase (DNA-dependent RNA polymerase, to be pedantic, though not necessarily always). The most common and favored polymerase to use today is T7, which, to be clear, is an enzyme I love. I’m a bacteriophage guy, and T7 polymerase is from the T7 bacteriophage. T7 polymerase does certain things well; for example, it has high processivity, good kcat [i.e., catalytic rates], and decent thermostability for an enzyme. But what we also know is, like all other bacteriophage-based polymerases, T7 doesn’t have high fidelity. So, the longer your reaction goes, the more likely your T7 polymerase is to abort, leading to truncated RNA sequences and lower IVT efficiency. These limitations are not necessarily a ‘bug;’ this is biologically how polymerases are supposed to work. Bacteriophage replication is error-happy to act as an evolutionary diversifying mechanism. But long story short, polymerases like T7 are good for what they’re good for — but they’re not good for anything else.

ARW: How are you seeing the industry respond to some of these limitations? Are we seeing more exploration of and/or production of alternative IVT enzymes?

Koeris: For sure, there are a handful of companies working on developing new enzymes or improving current T7 polymerase. But I think the challenge we face is that the commercial success of the process used for the COVID vaccines has been regulatorily derisked. As an industry, we tend to stick with what we know and, in this case, that means we’re predominantly relying on the as-yet limited processes and/or materials that have withstood the regulatory test.

ARW: What would “better IVT enzymes” mean/look like to you, and/or what benefits would these unlock for our molecules?

Koeris: We are going to need enzymes that have a higher fidelity and can do longer reads. The constructs for some of the next-generation RNA products, like self-amplifying or trans-amplifying RNA, are only getting bigger. However, if you look at the performance of enzymes over time, they don’t typically become gradually less effective; their activity reaches a certain threshold and then steeply plummets. So, the longer the RNA sequence, the greater the risk we face of our enzymes ‘shorting out’ before our sequence is completed.

ARW: Beyond exploring alternative constructs and raw materials that can help with next-gen RNA production, what other scientific advancements will be essential to enhance the overall performance of our RNA constructs?

Koeris: There’s still a lot of exploration that can be done on the nucleotide side. We already incorporate pseudouridine to enhance the stability of our mRNA. But there’s a whole zoo of modified nucleotides, many of which come from the phage world. In general, phages are exceptionally good at modifying their nucleotides to escape from various anti-bacteriophage mechanisms. But of course, the synthetic exploration of the chemical space knows no bounds, and here, too, generative AI can also provide enormous expansion of the space to explore. There are some folks working on quantifying nucleotides chemically today. But the question then becomes: How do we incorporate these nucleotides? Do we incorporate them specifically? Do we incorporate them stoichiometrically? Or do we incorporate them randomly? We haven’t spent a significant amount of time exploring these different approaches to determine which is best, and why. These questions will likely be asked by the FDA eventually — you’ve heard it here first!

ARW: Any ideas on what it will take for us to start answering these questions?

Koeris: An interesting gap I’ve observed is that nucleotide chemists typically don’t talk to nucleotide biologists. It’s not that they’re avoiding each other or not talking to each other at all, per se; the connection between these two fields is just not that common on an academic level. Part of this, too, is because there are just more scientists out there working on mRNA than there are nucleotide chemists. But overall, I see more opportunities for greater communication and collaboration between nucleotide chemists and nucleotide biologists.

ARW: Currently, pseudouridine is the only modification that is the “go-to” in the space. You mentioned the importance of exploring the chemical space more thoroughly; what benefits might this offer our products?

Koeris: You have your four base pairs of mRNA — A, U, G, & C— the base-pairing behavior of which informs the mRNA’s translation into a protein. But there are also the chemical properties of a nucleic acid which we’re also striving to manage better today, including stability and immunogenicity. By exploring our RNA’s chemical properties better, we can improve the function and utility of nucleic acids.

Modifying nucleic acids could also facilitate endosomal escape, which is in and of itself a complicated mechanism we often link predominantly to the LNP. But I think it’s necessary to consider the nucleic acid as a critical component in this series of biological events, as well.