Michelle Lynn Hall

When Michelle Lynn Hall was looking for a new job in 2016, Moderna wasn’t an obvious first choice. Hall, a computational chemist, was well versed in quantum mechanical simulations and atomic-level molecular dynamics—techniques useful in small-molecule and protein-based drug discovery.

Moderna had plans to do something very different: make drugs out of messenger RNA, the intermediate molecule between our genes and our proteins. The Cambridge, Massachusetts–based company often describes mRNA as software for cells, a molecular instruction manual for making therapeutic proteins that fight infections, combat cancer, and treat rare diseases. “It sounded like science fiction,” Hall says. “And that was the draw.”

But making mRNA therapy a reality is not so straightforward. mRNA is a large, unwieldy molecule. Predicting how its thousands of nucleotide building blocks will interact with different parts of cells—such as the machines that translate the mRNA code or the immune sentinels that spot and destroy foreign mRNA—is a headache. To help solve that problem, Iain McFadyen, Moderna’s head of computational sciences, wanted to hire the company’s first computational chemist. He needed someone undaunted by uncharted territory, and Hall fit the bill.

As a graduate student, Hall made a last-minute decision to join a computational chemistry lab even though she arrived at Columbia University intent on studying organic chemistry. When Hall’s postdoc adviser at Novartis gave her a list of potential projects to work on, she picked one that she knew nothing about. And as a researcher at Schrödinger, a chemical software company, Hall used machine learning to analyze gene expression data for drug discovery—a task well outside the firm’s expertise.

Hall hardly knew a thing about mRNA when McFadyen interviewed her, but he was impressed with her experience and fearlessness. “She was the only one I considered for the position,” he says.

Hall immediately set out to model mRNA’s structure and function, and her work has helped Moderna’s scientists design better molecules. In the process, she learned about another, even bigger challenge that could make or break mRNA therapy’s ability to live up to its full potential: the delivery dilemma.

To get mRNA into cells of the body, Moderna scientists were packaging it in fatty spheres called lipid nanoparticles. Lipid nanoparticles aren’t new. Other companies have used them to deliver small interfering RNAs, short nucleotide strands that can block the production of problematic proteins. The lack of safe and effective nanoparticles almost derailed the development of those therapies. Moderna doesn’t want the same thing to happen to mRNA therapy, and it needs new nanoparticles that can hold the comparatively giant mRNA molecules.

Each of Moderna’s lipid nanoparticles is a collection of about 60,000 lipid and cholesterol molecules. But scientists still don’t understand how those molecules coalesce into a structure that encapsulates mRNA. Hall thought that a computational model could help answer that question.

At Moderna’s annual science day in May, Hall presented the first results of that project. She used an approach called coarse-grained molecular dynamics to study how those thousands of nanoparticle parts interact. She discovered that the first 1.3 µs of the reaction are crucial for encapsulating mRNA.

Those insights will help Moderna’s lab chemists create new recipes and protocols for making nanoparticles that better package and deliver their mRNA cargo. The firm is still years away from getting its first drug approved—and there’s no guarantee that it will happen. But if it does, Hall will likely have played a vital role in transforming mRNA therapy from the realm of science fiction to reality.