The story of mRNA vaccines is usually told as a story of a single year. In December 2020, less than twelve months after the first cases of a new respiratory illness appeared in Wuhan, the United States authorized a vaccine for emergency use. Then the United Kingdom did. Then most of the world followed. By the end of 2021, four and a half billion doses had been delivered. It was the fastest, broadest medical deployment in human history.
This is the version of the story that gets retold at conferences. It is true, but it is also the wrong story. The work that made 2020 possible was done between roughly 1989 and 2005, by a small number of researchers nobody was paying particular attention to, on a problem most of their colleagues thought was a dead end.
The forty-year throat clear
The idea that you could put genetic instructions into a cell, have the cell make a protein, and use that protein to provoke an immune response — that idea was first written down in the early 1960s. It took until 1990 for anyone to demonstrate it in a mouse. It took another decade after that for anyone to do it in a way that did not immediately kill the mouse.
The problem was that messenger RNA, the molecule you would need to deliver, is among the most fragile substances biology produces. It is so fragile that the human body treats any free-floating mRNA as evidence of viral infection, and responds by setting the surrounding tissue on fire. For thirty years, every attempt to use synthetic mRNA as a drug ran into the same wall: the inflammation was worse than the disease.
"We knew the molecule worked. The cells made the protein. The only problem was that the patient died."
Karikó's contribution, made over fifteen years at the University of Pennsylvania under conditions of essentially total professional indifference, was a small chemical substitution. By replacing one of the four bases in synthetic mRNA — uridine — with a modified version called pseudouridine, the molecule became invisible to the immune system's tripwires. The cells still translated it. The tissue did not catch fire. A drug class became possible.
What it actually was
It is worth being precise about what happened in 2020. The platform was ready. The chemistry was solved. What was missing was a target — a specific viral protein worth aiming at — and the willingness, in a moment of public emergency, to spend the money to find out whether the platform would work in millions of people.
Both arrived at once. Within forty-eight hours of the SARS-CoV-2 genome being published, both BioNTech and Moderna had a candidate sequence. Within six weeks, the first vaccines were in vials. The compressed timeline that the public found so miraculous was the timeline of distribution, not the timeline of invention. The invention had taken a generation.
What comes next
The interesting question is no longer whether mRNA vaccines work. They demonstrably do. The question is what else you can do with a platform that can produce a working immunogen against essentially any target in under a month. Influenza vaccines built this way are already in trials and show roughly twice the efficacy of the traditional egg-based versions. Personalized cancer vaccines — trained on the specific mutations in a single patient's tumor — are showing complete responses in early melanoma trials. RSV, malaria, tuberculosis, and HIV are all targets the platform may finally crack.
The deeper change is institutional. For most of the twentieth century, a new vaccine was a decade-long undertaking by a handful of large pharmaceutical companies, each operating its own bespoke biological pipeline. The mRNA platform replaces the biology with chemistry — and chemistry is a thing small labs can do. The number of organizations capable of producing a clinical-grade vaccine has gone, in the space of five years, from perhaps a dozen worldwide to several hundred.
This is a different kind of biotechnology than we have had before. It is fast, cheap, modular, and within reach of countries and institutions that have never had it. Whether the regulatory and manufacturing infrastructure can keep up — whether we can move at the speed the molecule now permits — is the open question of the decade. The molecule, finally, is no longer the bottleneck.