To hear Tony Hunter tell the story, his discovery of tyrosine phosphorylation was nothing more than a happy accident. It was 1979, and researchers had known for decades that protein kinases were involved in regulating cell growth, proliferation, and metabolism. But, as far as everyone thought, kinases phosphorylated their target proteins only on serine or threonine residues, the only modified amino acids that had been detected.
Enter Tony Hunter. An assistant professor at the Salk Institute for Biological Studies, Hunter was studying how certain tumor viruses - in particular, polyomavirus and the Rous sarcoma virus - transform human cells. Although researchers had identified the key proteins involved, that is, middle T antigen for polyomavirus and Src for Rous, they weren't sure what those proteins did. "Two groups had reported a kinase activity associated with the viral Src protein," recalls Hunter: Marc Collett and Ray Erikson, then at the University of Colorado Medical Center, and Mike Bishop and Harold Varmus, then at the University of California, San Francisco. "They said it phosphorylates threonine. So we looked to see whether middle T also had an associated protein kinase activity, and we were delighted to find that it did." Indeed, middle T antigen was able to phosphorylate itself.
The next step was to determine the identity of the amino acid it modified. The experiment was fairly routine. Hunter ran the P32-labeled phosphorylated protein on a gel. He excised the middle T band, eluted the protein, and performed a partial acid hydrolysis to release the phosphorylated amino acids. He then spotted this hydrolysate onto a thin-layer plate along with markers for phosphoserine and phosphothreonine and ran the electrophoresis. If the phosphorylated amino acid released from the protein comigrated with phosphoserine, then middle T must phosphorylate serine. If it comigrated with phosphothreonine, then middle T would have threonine kinase activity.
"Much to my amazement, the P32 ran between phosphoserine and phosphothreonine," says Hunter. He repeated the experiment and got the same strange result. Then Hunter remembered a third amino acid with a free hydroxyl group, tyrosine, which could potentially be phosphorylated. "I called around and no one had ever heard of a tyrosine kinase," says Hunter, but he thought he should check it out. So he synthesized some phosphotyrosine and spotted that onto a plate along with his other markers and sample. "Sure enough, phosphotyrosine ran in this space between phosphothreonine and phosphoserine," he says. "But more importantly, it comigrated with the labeled compound I was generating from middle T."
The reason that Hunter saw phosphotyrosine where no one else had, he says, was because he was "too lazy to make up fresh buffer" before doing the experiment. To electrophoretically separate phosphothreonine from phosphoserine, Hunter, like everyone else in the field, had been using a buffer of pH 1.9. At that pH, "phosphotyrosine and phosphothreonine run on top of one another," he says. But Hunter's buffer wasn't actually pH 1.9. "Because the buffer was old and had been used many times," he says, its pH had dropped to 1.7, allowing Hunter to resolve, for the first time, phosphothreonine and phosphotyrosine. Indeed, he then demonstrated that the viral Src protein was also a tyrosine kinase, and not a threonine kinase as had previously been reported.
Crediting the discovery entirely to happenstance might be erring on the side of modesty, says Tomas Mustelin of the Burnham Institute in La Jolla, Calif. "The real genius lies in the fact that Tony realized that an accident was as significant as an intentionally discovered result," he says. "Most people would have said, 'I screwed up, let me fix the problem.' But Tony was smart enough to say, 'If the standard and the label did not comigrate, maybe they aren't the same.'"