All of this is how mRNA vaccines should work in theory. But no one on Earth, until last week, knew whether mRNA vaccines actually do work in humans for COVID-19. Although scientists had prototyped other mRNA vaccines before the pandemic, the technology was still new. None had been put through the paces of a large clinical trial. And the human immune system is notoriously complicated and unpredictable. Immunology is, as my colleague Ed Yong has written, where intuition goes to die. Vaccines can even make diseases more severe, rather than less. The data from these large clinical trials from Pfizer/BioNTech and Moderna are the first, real-world proof that mRNA vaccines protect against disease as expected. The hope, in the many years when mRNA vaccine research flew under the radar, was that the technology would deliver results quickly in a pandemic. And now it has.

“What a relief,” says Barney Graham, a virologist at the National Institutes of Health, who helped design the spike protein for the Moderna vaccine. “You can make thousands of decisions, and thousands of things have to go right for this to actually come out and work. You’re just worried that you have made some wrong turns along the way.” For Graham, this vaccine is a culmination of years of such decisions, long predating the discovery of the coronavirus that causes COVID-19. He and his collaborators had homed in on the importance of spike protein in another virus, called respiratory syncytial virus, and figured out how to make the protein more stable and thus suitable for vaccines. This modification appears in both Pfizer/BioNTech’s and Moderna’s vaccines, as well as other leading vaccine candidates.

The spectacular efficacy of these vaccines, should the preliminary data hold, likely also has to do with the choice of spike protein as vaccine target. On one hand, scientists were prepared for the spike protein, thanks to research like Graham’s. On the other hand, the coronavirus’s spike protein offered an opening. Three separate components of the immune system—antibodies, helper cells, and killer T cells—all respond to the spike protein, which isn’t the case with most viruses.

In this, we were lucky. “It’s the three punches,” says Alessandro Sette.