Modeling Macromolecular Complexes: A Journey Across Scales - Modeling in Computational Biology and Biomedicine - page 25

1. The channel cylinder, containing proteins having unstructured regions (filaments)
that regulate active transport;
2. The adapter cylinder, which occupies an intermediate position between channel
and scaffold;
3. The coat cylinder, which defines the scaffold of the NPC;
4. The pore membrane cylinder, anchoring the NPC in the nuclear membrane.

Making a precise statement on the stoichiometry of the protein types in the NPC
is difficult per se, as the proteins present indeed change along the life cycle of the
cell. During cell division, with concomitant nuclear division, selected NPC proteins
diffuse back into the cytoplasm and join the new nuclei at a later stage.
This plasticity, together with the large number of proteins involved and their
inherent flexibility, makes the reconstruction of such a molecular assembly a highly
challenging endeavor.
As of now, two types of structural results are available for the NPC, as illustrated
on Fig. 1.8 . On the one hand, atomic resolution structures of monomers and dimers
have been resolved, from which plausible sub-complexes of the NPC have been
modeled. One example is the Y-complex, an heptameric assembly found in the coat
cylinder, making up the scaffold of the NPC [ 61 ]. Another complex reconstructed
at atomic resolution is the so-called T -complex, whose unstructured filaments are
believed to participate in active transport across the NPC [ 60 ]. On the other hand,
putative global models of the NPC in its entirety have also been reconstructed
by integrating data providing complementary information [ 3 , 4 ]. A variety of data
can be used in these models, and the following three deserve a special mention.
First, the overall shape of the assembly is typically inferred from cryo-electron
microscopy. Second, the 3D location of particular proteins can be captured thanks to
immuno electron-microscopy, a process which consists of tracking under an electron
microscopy proteins coated with antibodies that have been conjugated with gold
particles. Third, particular interactions between pairs of proteins can be inferred
from proteomics experiments, in particular Tandem Affinity Purification (TAP) data.
Such data are usually noisy and/or ambiguous, as we shall discuss in an example
in the next section.

1.3.1.2 Low Resolution and Ambiguous Data: The Example of TAP data

As discussed in Sect. 1.1.1 , structural information for assemblies can be obtained
from cryoEM. Such data provide information at the scale of the assembly, yet
because of their noisy nature complementary information is needed in order to
exploit them. Of particular interest are proteomics data, which provide information
on the interaction between protein types, which we illustrate with Tandem Affinity
Purification (TAP) data.
TAP experiments give access to all types of proteins found in all complexes that
contain a prescribed protein type, say R . More precisely, the method consists of
the following steps. First, a fusion protein is created by modifying the gene for