One gene, multiple proteins, new clues to rare disease

Around 25 million Americans have rare genetic diseases. Many of them have spent years searching for answers—getting tested, getting nowhere, living with symptoms that don't quite fit any diagnosis in the medical literature. Up to 70% still lack a clear genetic explanation. But researchers at Whitehead Institute may have found why doctors have been looking in the wrong place.

The assumption has always been straightforward: one gene makes one protein. If a mutation doesn't break that protein, the gene gets cleared. Case closed. Except it's not. A new paper in Molecular Cell shows that most genes actually code for multiple proteins, not just one. A single mutation might leave the "main" protein intact while breaking a hidden version that cells were quietly making all along.

This changes everything about how we should be reading the genetic code.

How One Gene Becomes Many Proteins

Cells are more creative than the textbook suggests. During protein production, the cellular machinery can choose where to start reading the genetic instructions. Skip the first starting point and pick up at a second one, and you get a shorter protein. Start reading earlier, at a section that looks like a starting point, and you get a longer version. Neither is a mistake—evolution has preserved this ability across species because it's useful. Different versions of the same protein can be sent to different parts of the cell: one to the mitochondria (the cell's power plant), another to the nucleus (where DNA lives). Same gene, different jobs.

But if a mutation only breaks one version of that protein, most diagnostic tools won't catch it. The other version keeps working. The gene looks fine. The patient stays undiagnosed.

The Two Patients Who Didn't Fit

Researchers found a test case in a rare blood disorder called SIFD (syndromic immunodeficiency with facial dysmorphism). Most patients with SIFD have mutations that damage both versions of the protein. But two patients broke the pattern. Their mutations only affected one version each.

The results were striking. One patient was missing the mitochondrial version of the protein. She developed anemia—the hallmark of SIFD—but her development was normal and her immune system worked fine. The other patient was missing the nuclear version. He had immune problems but no anemia and no developmental delays. Same disease, completely different bodies, because different protein versions were affected.

This wasn't just a curiosity. It meant clinicians had been missing a whole category of genetic explanations because they weren't looking for mutations that only broke part of the protein story.

What Comes Next

The team is building a tool called SwissIsoform to help doctors spot mutations affecting specific protein versions. It's a practical step toward a larger shift: training clinicians to think about genes as multi-protein factories, not single-product lines. The work could reshape how rare diseases are diagnosed and eventually how gene therapies are designed—because if you're going to fix a gene, you need to know which version of the protein actually needs fixing.

For researchers involved, there's a particular satisfaction in knowing this isn't just theory. Somewhere, a patient who's been searching for answers might finally get one.