Order, Chaos, And The mRNA Black Box: Takeaways From The USP/AMM Forum

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

It sounds like the start of a bad joke: A poet walks into an mRNA analytical forum.

But this was exactly what happened in mid-April when I had the opportunity to attend the USP and AMM joint workshop, “End-to-End Analytical Solutions for mRNA Therapeutics: Supporting Manufacturing & CMC Needs.”

For two days, my colleague Michael Soloway and I were privy to presentations on the myriad advancements being made with novel methods, impurity analysis (namely dsRNA), and the establishment of platform methods.

Those of you who are regular readers of Advancing RNA may remember that Britney Spears played a starring role in my takeaway article from last year’s forum. During last year’s workshop, we heard presentation after presentation on the often-painstaking work we’re doing to better understand the limitations of our methods and improve their repeatability (cue Britney’s “Oops I did it again”), particularly to establish much needed platform methods.

To be clear, this year’s forum was just as focused on the intricacies of applying/using the methods. But what was equally if not more present in this year’s discussion was our increasing acknowledgement that our understanding of mRNA and its biological mechanisms remains quite messy.

We often talk about mRNA in simplistic terms — as if it’s a puzzle we’ve solved analytically. But outside of ensuring that our product has all the necessary physical attributes (cap/tail/ORF/UTRs), we’re also increasingly aware that we’re sending our products into the “wild west” of the cell, and we are dealing with a relatively complicated biological cascade. As consultant and Advancing RNA editorial board member Khaled Yamout succinctly described it, “There is a multi-step pathway from delivery to translation.” In other words, we ultimately need a variety of things to go right inside the cell for our product to be translated into the proper protein and to have a therapeutic effect. We also need methods capable of qualifying and quantifying these mechanisms.

“Interrogating biology is hard to do,” Eclipse Bio’s Wayne Doyle furthered. “In practice, mRNA is a biological system, so we need assays that are more suited to the biological processes of what’s actually happening within the cell to better understand it.”

Fortunately, there were a few discussions throughout the forum that demonstrated our ability to turn to methods, such as next-gen sequencing, to better understand the interplay between our products and the environs of a cell.

This was particularly true for our inquiries into the oft-discussed dsRNA. As this forum demonstrated, we’re no longer interested in just understanding how dsRNA is formed and how to quantify it; we’re also trying to understand why and to what extent dsRNA is biologically impactful. As Doyle furthered, sequencing has not only taught us more about the formation of dsRNA species and how to optimize our processes to avoid dsRNA formation, but it has also given us much more insight into how dsRNA affects cellular behavior — namely, how dsRNA stymies ribosome performance and limits overall protein translation.

But dsRNA discussions aside, what became clear throughout the two days is that we’re still finding it hard to answer the “so what?” behind many of our measurements of mRNA’s structure.

“We did all this work on physical characterization and a lot of it is really, really phenomenal work because we brought this modality leaps and bounds,” Yamout added. “But what does this physical characterization say about translation and function? We need to make that bridge. That's what we're missing.”

“We can't figure out which structures are relevant and which ones are not,” Cytiva’s Adam Crowe furthered. “Even just adding a bit of salt can change the structure of an mRNA. I think a lot of those structural differences are not relevant, but I do think there are central ones.”

As he went on to explain, it’s not unusual to find ourselves in a situation where the same mRNA construct from different vendors (with seemingly water-tight panels of CQAs) reveals different potencies. This raises the controversial question: Are we missing some key CQAs? There’s a fine line between order and chaos, and I daresay Crowe’s question takes us a little bit closer to chaos.

I acknowledge that it’s strange to bring the word “chaos” into an article about analytical development, given that analytical development is a critical means of establishing control and order. But running undercurrent to all the meticulous presentations at the USP/AMM Forum was this acknowledgement that, despite our characterization thus far, mRNA remains a black box in many ways. As an industry, we have to get more comfortable with (and be more open to) the messiness of answering the all-important question, “So what?”

The journey toward answering this question means a couple of different things. For Yamout, this means linking physical characterization to functional characterization in the cell — a task we’re not doing enough of today. We need more data on the function of our product in the cell (and about the protein it encodes) to guide our molecular optimization — not the other way around. (For more on this be sure to check out this previous Advancing RNA article I penned with Yamout!)

We also have to get much more comfortable with ambiguity and messiness in our analytical data. If there was one quote that I feel sums up my biggest takeaway from this forum, it would be this one from Crowe who emphasized how important it is to interrogate our mRNA from every possible angle — beautiful and orderly data be damned.

“There's a bias towards having nice looking data,” he concluded. “The only way you get that data is by denaturing the hell out of our mRNA. I think we're missing a lot of information by trying to get that single peak, that single beautiful result, where the reality is, if you don't do that, mRNA is a complex mix of different structures and different profiles. I think we have to start embracing that messiness. We need to look at mRNA in its fullest state and stop trying to make it the prettiest picture we can make.”