Karikó and Weissman modified mRNA's bases so the body stopped fighting it

By the early 1990s, mRNA was a seductive therapeutic idea: in-vitro transcription could produce a template for any protein without cell culture. But lab-made mRNA was unstable, hard to deliver, and inflammatory — dendritic cells treated it as a foreign substance and released signaling molecules that made any therapy self-defeating. Katalin Karikó, then an assistant professor at the University of Pennsylvania, kept pursuing mRNA therapy despite failing to convince funders, and a collaboration with the immunologist Drew Weissman took up how different RNA types interact with the immune system.

They noticed that mRNA from mammalian cells did not trigger the reaction that lab-made mRNA did, and knew that mammalian RNA bases are frequently chemically modified while in-vitro transcripts are not. Testing mRNA variants with different base alterations, they found the inflammatory response almost abolished when modifications were included — a result published in 2005, fifteen years before the COVID-19 pandemic. Further studies in 2008 and 2010 showed modified mRNA also markedly increased protein production.

The 2008 Molecular Therapy paper quantified the trade in mice: mRNA containing pseudouridine showed higher translational capacity in cells, lysates and after intravenous injection, persisted detectably in the spleen at 1, 4 and 24 hours, and — unlike unmodified mRNA — did not raise serum interferon-α even at higher doses. Because mammalian RNA abundantly carries modified nucleosides, the sensors that guard against bacterial RNA simply stopped responding.

Make lab-made RNA look like the body's own

The immune reaction was a recognition problem, not a dosage problem: sensors like TLR3/7/8 and RIG-I treat unmodified RNA as bacterial or viral, so no amount of purification would fix it.

Pseudouridine was not a new synthetic chemical but a natural modification abundant in mammalian RNA, so it made synthetic transcripts mimic the body's own rather than evade detection artificially.

The same modification that silenced the alarm also improved the drug: higher translational capacity and stability, so a safer molecule was also a more potent one.

The platform generalised — once any transcript could be delivered without inflammation, only the encoded protein needed to change, letting two spike-protein vaccines be built at pandemic speed.

The obstacle was not delivery or dose but identity: the molecule announced itself as an invader. Borrowing a chemical signature the body tolerates turned a rejected molecule into a platform.

Interest picked up around 2010 as companies including BioNTech and Moderna built on the finding. After COVID-19 emerged, two base-modified mRNA vaccines encoding the SARS-CoV-2 spike protein were developed at record speed, reported protective effects around 95%, and were approved as early as December 2020; more than 13 billion COVID-19 vaccine doses have since been given globally. The Royal Karolinska Institute awarded Karikó and Weissman the 2023 Nobel Prize in Physiology or Medicine for the base-modification discoveries.