Figure 1.

Similarities between COPII coats and nuclear pore coats. (A) Structure of a COPII coat driving vesicle budding from the ER membrane. The integral membrane protein (Sec12) marks the site of budding and recruits Sar1 to the membrane. Sar1 inserts itself in the outer leaflet of the membrane via an amphiphatic α-helix and imparts curvature to the membrane (Lee et al., 2005). Sar1 then recruits the Sec23–Sec24 coat complex, which traps membrane protein cargos using a large concave surface (Fromme et al., 2008). Another set of coat proteins (Sec13–Sec31) docks to Sec23–Sec24 and oligomerizes to create an outer cage structure (Fath et al., 2007) that helps Sar1 drive membrane fusion and vesicle scission (Lee et al., 2005). (B) Structure of the proposed nuclear pore membrane coat formed by Poms, Nups, and reticulons. See the main text for a detailed description. The topology of the Nup84 complex (as shown) is different than suggested in previous models (Alber et al., 2007; Hsia et al., 2007) but is consistent with the data used to generate them. In this alternate configuration, 16 concave-shaped Nup84 heptamers (each ∼8 nm wide × 40 nm long; Hsia et al., 2007) would stack laterally in an antiparallel arrangement (joined at the midbody) with their concave sides facing outwards toward the pore membrane. This would create an hourglass scaffold structure with a minimum inner diameter of 40 nm, which is consistent with the size limit for facilitated transport across NPC. Interestingly, this layout could physically isolate the pore membrane region from the central conduit region. (C) Diagram depicting what happens to cells when supply lines of Poms and Nups that provide the foundation for building new NPCs are disrupted. Without Pom152 and Pom34 (scenario 1) or Nup53 and Nup59 (scenario 2), Ndc1 and cytoplasmic facing nucleoporins (e.g., Nup82 and associated Nups; Alber et al., 2007) are not secured at new sites of NPC assembly and can wander off path to peripheral ER membranes (Onishchenko et al., 2009). Without the reticulons Rtn1 and Yop1 (scenario 3), the curved membranes at sites of new NPC assembly would be unstable, leading to the accumulation of NPC-like intermediates at the nuclear envelope (Dawson et al., 2009). Without the Nup170/Nup157 homologues (scenario 4; Makio et al., 2009) or in cells depleted of Nup170 but overexpressing its C terminus (scenario 5; Flemming et al., 2009), the Nup53-Ndc1-Pom34-Pom152 complex is unstable and unable to serve as a foundation for building new NPCs. A normal intermediate of NPC assembly depicting a later stage is shown for comparison (scenario 6). A more mature NPC is shown in B.