Although Schekman is still studying membrane trafficking in yeast, about half of his lab has turned to mammalian cells. His students and postdocs are probing possible trafficking defects in familial forms of Alzheimer's disease and exploring illnesses known to stem from malfunctioning secretory genes.
For example, the molecular defect underlying craniolenticulosutural dysplasia (CLSD), a disease that affects the facial skeleton, was a mystery until recently. In 2006, researchers established that the condition arises from a marred version of Sec23, a protein that Schekman discovered in yeast. Now his lab has discerned how this miscreant molecule is causing trouble: the ER cannot export proteins properly because flawed Sec23 cannot bind and recruit two other Sec proteins that normally shape new vesicles and help them pinch off.
In the same issue of Developmental Cell where Schekman reported these findings, in November 2007, Jonathan Goldberg, an HHMI investigator at Memorial Sloan-Kettering Cancer Center, published the crystal structure of those two Sec proteins, showing that they snuggle up next to the amino acid in Sec23 that is altered in CLSD.
The observation that these proteins embrace this specific amino acid suggests that changing it would disrupt the interaction, just as Schekman revealed through his experiments. “It's just too good to be true,” he says of the corroborating evidence.
But Schekman is not turning ice buckets into footrests yet. He's moving on to another cellular mystery: how peroxisomes are formed. These membrane-bound organelles house enzymes that detoxify and break down various molecules within cells. Defects in them can lead to serious diseases (see “Parsing Peroxisomes,”). Scientists had assumed that peroxisomes arise by duplicating themselves rather than by budding off from other membrane-bound structures: no one had detected peroxisomal proteins in other membranes and peroxisomes form in the absence of SEC genes.
In 2005, however, researchers discovered that one of the two proteins known to be essential for peroxisome formation clusters in the ER and pinches off in a vesicle that becomes a brand new peroxisome. Because SEC genes do not contribute to this process, the observation implies the existence of a membrane-trafficking pathway that depends on mostly unidentified genes. Schekman is devising ways to uncloak members of this possible transport network.
The challenge poses “another opportunity to discover the rules that govern protein segregation,” says Schekman. If SEC-related vesicles are not involved, “what the hell is? We're going to go back and think more clearly, I hope, about how to find the right genes.”