For most of us, the word "switch" conjures something simple: on or off, up or down. But researchers at North Carolina State University have discovered that the proteins controlling our genetic code are far more nuanced than any household light switch—and that subtlety may hold the key to better medicines and smarter biological computers.
In a study published in the journal iScience, Albert Keung, an associate professor of chemical and biomolecular engineering at NC State, led a team that tested how 87 different proteins interact with a single gene from yeast. The results upended the assumption that these proteins—collectively known as the epigenome—merely turn genes on or off.
"One protein may turn the gene on quickly; a second protein may take slightly longer to turn the gene on—but then keep it on for a long time; and a third protein might have a long time delay before turning the gene on, at which point it spikes up quickly and then turns off right away," explained Leandra Caywood, co-first author and recent Ph.D. graduate from NC State.
The team exposed the gene to each protein in approximately 100 yeast cells, using light to control protein binding and microscopy to measure gene expression in real time over 12 hours. What they found was remarkable: every protein produced its own distinct pattern of gene behavior—akin to different musical instruments playing the same note but with entirely different rhythms and tones.
"The big finding here was that each protein produced a uniquely patterned response of gene expression from the gene," said Keung, who is also the Goodnight Distinguished Scholar in Innovation in Biotechnology and Biomolecular Engineering. "The proteins are far more than an on/off switch."
Some proteins produced remarkably consistent results across all tested cells, while others created what scientists call "noise"—widely varying responses from cell to cell. This variety, the researchers realized, isn't chaos; it's sophistication. When the team analyzed the patterns, they found a strong connection between what scientists already knew about each protein's biological function and the expression patterns it produced. Proteins known to recruit polymerase, for instance, tended to share similar behavioral signatures.
Perhaps most surprisingly, a relatively simple computational model—a three-state model with positive feedback—captured all the complex data the researchers had collected. "We initially thought this would be difficult," Keung noted, "because there were so many different gene expression patterns."
The implications ripple outward. Understanding these precise protein-gene dynamics could help scientists design more targeted cellular therapies, engineer biological systems with greater precision, and even advance the emerging field of biological computing. It's a reminder that the machinery of life operates with an elegance that continues to surprise even the scientists studying it.