On the Mission Bay campus of the University of California, San Francisco (UCSF), Peter Walter's corner office is distinctive for its tall, elegant cactus-like plants—and its poetry-quoting African grey parrot. After months of effort, Walter has trained the parrot, named Beaker, to badger lab members on what's most important: “We need more data.”

That's a common refrain from research leaders, but “in this case, Peter's got a parrot to say it,” says postdoctoral researcher Jonathan Lin, laughing.

In keeping with Beaker's command, Walter's team has collected convincing data to explain exactly how the endoplasmic reticulum (ER), a maze-like compartment that fans out around the nucleus inside the cell, can determine whether a cell lives or dies. The ER serves as a cellular factory where newly synthesized proteins are folded into their proper structure. A protein must be folded correctly to do its job.

“Only the perfectly made proteins pass quality control,” says Walter, an HHMI investigator. Proteins that fail to attain the right conformation are degraded before they can cause cellular dysfunction and disease.

If the ER machinery is insufficient or defective, however, unfolded proteins pile up. Fifteen years ago, in studies of yeast, Walter and his colleagues discovered that the ER copes with the stress of such overload by triggering a set of biochemical reactions, known as the unfolded protein response, or UPR. Later work by various research groups uncovered a similar mechanism in mammalian cells, including human cells: three enzymes, molecular sensors called IRE1, ATF6, and PERK, detect the glut of unfolded proteins. They then activate various genes that expand the ER and step up its folding capacity, reduce the synthesis of new proteins, and crank up the protein degradation process.

Those protective measures bring the system back into balance. “It's a feedback loop that adjusts supply to demand,” Walter says. Yet, paradoxically, if the ER cannot regain equilibrium, the UPR prompts the cell to commit suicide. In a study published in Science last November, Lin, Walter, HHMI investigator Kevan Shokat, and colleagues explored how the same signaling pathways could cause such diametrically opposite fates.

The team exposed cultured human cells to drugs that prevent proteins from folding and eventually cause cell death. Over 24 to 30 hours, the researchers measured the activity generated by the UPR. Initially, all three pathways rapidly turned on, but results unexpectedly showed IRE1 shutting off after about 8 hours, around the time when cells began to deteriorate. ATF6 activity followed a similar pattern. By contrast, responses triggered by PERK—including production of a protein that promotes cell suicide—stayed on the whole time.

Illustration: Andy Smith

	PAGE 1 OF 2	Continue