How proteins cross the Nuclear Pore Complex
Over this past summer I saw Dirk Görlich give a talk about how the multitude of FG repeats found within the nuclear pore complex (NPC), form a gel like matrix. This "elastic hydrogel" acts as the major barrier within the NPC. Although the gel can prevent the passage of most large molecules (>30kD), it is permeable to nuclear transport receptors (NTRs). Note that all this "story" was published in the November 3rd edition of Science Magazine (link). In that paper there's a nice diagram in the that explains it all:
An "FG repeat" is a long stretch of amino acids that form non-covalent cross linking type interactions to form a gel like substance within the pore. FG repeats are found in many NPC proteins, and each of these proteins contains many of these motifs. (Shown here are the FG repeats found in Nsp1, a single NPC protein). Proteins that are too big to diffuse within the holes of the matrix are barred from crossing the gel. NTRs on the other hand can displace the FG-FG interactions by forming FG-NTR interactions and thus participating in the gel's scaffold. This property allows the NTRs (and associated proteins) to cross the matrix while maintaining the integrity of the gel scaffold (see cartoon above).
Gorlich's lab has even been able to form this gel in the lab. They show that if enough of the FG repeats are mutated, then the interactions are broken and the gel can't form. These mutations also make the NPC more leaky in yeast.
So it turns out that the core of the NPC is like a gummy bear that can absorb like molecules and exclude everything else.
Steffen Frey, Ralf P. Richter, Dirk Görlich
FG-Rich Repeats of Nuclear Pore Proteins Form a Three-Dimensional Meshwork with Hydrogel-Like Properties
Science (06) 314:815-817