THE EMBO JOURNAL

Transport into and out of the nucleus is fundamental for many aspects of cellular function. Cargo molecules are transported through nuclear pore complexes (NPCs). Small molecules can pass by diffusion, but there is a permeability barrier for larger molecules, with a relative molecular mass of >20–40K, which permits transport of only selective cargo with the help of specific nuclear transport receptors. However, the nature of this barrier and the precise molecular mechanism for selective transport are largely unknown. In a recent paper (EMBO J. 20, 1320–1330, 2001), Ribbeck and Görlich provide an intriguing model, the 'selective-phase hypothesis' which integrates both aspect of translocation.

In a system optimised to avoid other rate-limiting factors, the authors measured the translocation capacity of NPCs using various model substrates, and observed maximal translocation rates on the order of 103 translocations per second, which is much higher than previously thought. For a homodimer of the nuclear transport receptor NTF2, for example, this rate is relatively close to the free diffusion rate through a hypothetical NPC lacking a permeability barrier. In contrast, translocation of green fluorescent protein (GFP), which has a similar molecular size, is about 100-fold slower.

How can this difference be explained? It is known that translocation facilitated by nuclear transport receptors requires that the receptors interact with phenylalanine (Phe)-rich repeats found in high numbers on many NPC components. The concentration of these repeats in the nuclear pore is estimated to be as high as 50 mM. Ribbeck and Görlich propose that these repeats (light blue circles) form weak hydrophobic interactions with each other and constitute the molecular equivalent of the so-called central plug, a low electron density structure that fills electron micrographs of nuclear pores. Thus a sieve-like structure is formed that allows the diffusion of small molecules but prevents the passage of larger ones. In this hypothesis, nuclear transport receptors (orange circle), through their ability to interact with these repeats (dark blue circles), partition into and pass through this meshwork. As the hydrophobic interactions between Phe-repeats are relatively weak, nuclear transport receptors would only require micromolar affinities to disrupt Phe-repeat interactions and thereby pass through the central plug. Such low affinity interactions and the corresponding high off-rates are consistent with the speed of nuclear transport.

Intriguingly, this model mirrors the concept of lipid bilayers, which also exhibit a permeability barrier and allow only the passage of molecules that can partition into the hydrophobic core of the bilayer. Although further work is required to provide experimental support for the 'selective-phase hypothesis', it is an exciting step forward in providing a model for both the molecular nature of the permeability barrier and for the selective ability of nuclear transport receptors to pass through NPCs.