The drop off in IRE1 appeared to be “a switch between the pro-survival versus the pro-death phases of the UPR,” says Lin. To test that hypothesis, he and Walter wanted to see what would happen if IRE1 did not power down. Fortunately, UCSF colleague Shokat gave them a “wonderful trick” to do just that, says Walter. Shokat used genetic methods in yeast to alter IRE1's structure so that the sensor could be selectively turned on by a designer drug (see sidebar).

Lin and Walter repeated their cell culture experiments, this time using human cells engineered with the mutant version of IRE1. Adding Shokat's drug artificially stimulated and sustained IRE1 levels in the cells—and substantially fewer of them died, confirming the researchers' theory that the enzyme was pivotal for cell survival.

Going a step further, the researchers examined developing eye cells in rats with retinitis pigmentosa. This inherited form of blindness results from degeneration of retinal cells that make misfolded light-sensing proteins. Those experiments revealed a downturn in IRE1 signaling, typical of cell suicide.

The study raises fresh questions: could future drugs be designed to enhance the UPR's protective responses or stave off overzealous cell suicide that occurs in this and other diseases—such as diabetes and Alzheimer's disease—in which cells die from protein-folding glitches? Lin is exploring that possibility in the blind rats.

Walter is investigating the other side of the coin. Could inhibiting the UPR's protective side within cancer cells, which must crank out many proteins to sustain rapid growth, put an end to a tumor's growth? Beaker the parrot's likely response is: “We need more data.”

		It's a tale of great chemistry between two HHMI investigators at the University of San Francisco: around 2000, Peter Walter talked to Kevan Shokat about devising a drug that would work in yeast to selectively suppress IRE1, which belongs to a vast family of enzymes called kinases. Shokat already had a designer compound in hand that blocked other kinases after he modified them to respond to it. Taking the same approach, he used genetic tinkering to slightly widen IRE1's active site, the pocket where energy-molecule ATP normally plugs in and activates the enzyme. Those changes permitted Shokat's drug to fit only into that mutated pocket, blocking ATP binding. Shokat expected this chemical-genetics strategy to shut off the enzyme, but the reverse happened. “We've done this with 100 kinases and IRE1 is the only one where the drug actually turned on the function of the kinase,” he says. “So that was an absolute surprise.” A surprise that proved helpful when Walter needed a persistent, rather than a suppressed, IRE1. —I.C.

	PAGE 2 OF 2	Back