A plug (green) stoppers the protein-conducting channel (seen from top).


Membrane-bound compartments present an immediate problem: proteins need to get into and across those membranes. Now, we have the first picture of how that process works, thanks to the structure determined by Bert van den Berg, William Clemons, Jr., Stephen Harrison, Tom Rapoport (Harvard Medical School, Boston, MA), and colleagues. They find that a single SecY complex from Methanococcus jannaschii forms an hourglass-shaped pore with a basal plug that probably swings out of the way to let nascent protein chains pass.

The complex was not an easy target: the structure came only after five years of experiments with proteins from ten different organisms. The effort was worthwhile. “In contrast to ion channels and other structures … where people had models before the structure was solved, in this case we really didn't have any ideas of how this would work,” says Rapoport. “We were shocked because we didn't expect the pore to be in one complex.”

Earlier EM experiments had suggested that a large pore formed in a gap between four associated complexes. But more recent EM data are consistent with this supposed pore being only an indentation, and the new structure clearly suggests a path for nascent proteins through a single complex.

Translocation begins, according to cross-linking data, when the signal sequence of the nascent protein inserts between transmembrane domains (TM) 2b and 7 of the main, channel-forming α-subunit. This probably has two effects. First, a plug formed by TM2a swings ∼22 Å out of the way, revealing a narrow, central constriction delimited by a ring of six hydrophobic residues. Second, the two pseudosymmetric halves of the channel are pried open a little to widen the pore ring. Further separation of the two halves should allow release of membrane-spanning domains of the nascent protein.

The plug and pore ring are Rapoport's candidates for forming the tight seal that prevents passage of molecules other than the translocating protein. This sealing function had previously been ascribed to either the ribosome in the cytoplasm or (in eukaryotes) BiP in the lumen. Now, Rapoport needs to check that his presumed pore region is where the polypeptide really goes, and that the channel works in single copy and using the proposed plug movement. In the longer term he hopes to see the channel in action with a ribosome or even a translocating polypeptide; but for now, he says, seeing the current structure “has been a dream for me.” ▪


van den Berg, B., et al. 2003. Nature. 10.1038/nature02218.