Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2016 Feb 18;14(2):e1002365.
doi: 10.1371/journal.pbio.1002365. eCollection 2016 Feb.

Interactome Mapping Reveals the Evolutionary History of the Nuclear Pore Complex

Free PMC article

Interactome Mapping Reveals the Evolutionary History of the Nuclear Pore Complex

Samson O Obado et al. PLoS Biol. .
Free PMC article


The nuclear pore complex (NPC) is responsible for nucleocytoplasmic transport and constitutes a hub for control of gene expression. The components of NPCs from several eukaryotic lineages have been determined, but only the yeast and vertebrate NPCs have been extensively characterized at the quaternary level. Significantly, recent evidence indicates that compositional similarity does not necessarily correspond to homologous architecture between NPCs from different taxa. To address this, we describe the interactome of the trypanosome NPC, a representative, highly divergent eukaryote. We identify numerous new NPC components and report an exhaustive interactome, allowing assignment of trypanosome nucleoporins to discrete NPC substructures. Remarkably, despite retaining similar protein composition, there are exceptional architectural dissimilarities between opisthokont (yeast and vertebrates) and excavate (trypanosomes) NPCs. Whilst elements of the inner core are conserved, numerous peripheral structures are highly divergent, perhaps reflecting requirements to interface with divergent nuclear and cytoplasmic functions. Moreover, the trypanosome NPC has almost complete nucleocytoplasmic symmetry, in contrast to the opisthokont NPC; this may reflect divergence in RNA export processes at the NPC cytoplasmic face, as we find evidence supporting Ran-dependent mRNA export in trypanosomes, similar to protein transport. We propose a model of stepwise acquisition of nucleocytoplasmic mechanistic complexity and demonstrate that detailed dissection of macromolecular complexes provides fuller understanding of evolutionary processes.

Conflict of interest statement

The authors have declared that no competing interests exist.


Fig 1
Fig 1. Affinity capture of the trypanosome NPC and identification of new Nups.
(A) Schematic of the eukaryotic phylogenetic tree, adapted from Field et al., 2014 [38], highlighting the close evolutionary distance between yeast and humans versus more divergent eukaryotes such as trypanosomes (Excavates). SAR and CCTH correspond to Stramenoplies, Apicomplexa, Rhizaria and Cryptophyta, Centrophelida, Telonimia, Haptophyta, respectively. FECA and LECA refer to the first and last eukaryotic common ancestors. (B) Using the green fluorescent protein (GFP)-tagged nuclear basket protein Nup110 (marked with a ‡), we affinity isolated structural components of the NPC (dark grey), FG repeat containing Nups (green), and specifically associated proteins (light grey), which include transport factors and the major trypanosome lamina protein NUP-1. Affinity isolates were resolved by SDS-PAGE and visualized by Coomassie staining. Protein bands were excised and identified by mass spectrometry (MS). We discovered five new nucleoporins (in bold); assignments are based on secondary structure prediction and localization, as well as multiple pullouts that indicate bona fide association with trypanosome NPC components. Putative nuclear envelope proteins, α/β tubulin, and known contaminants (immunoglobulin heavy chain, variant VHH, and light chains of polyclonal llama anti-GFP antibodies) are marked by asterisks. A comprehensive list of all proteins identified is shown in S1 Fig. A schematic of the NPC is shown to highlight the architecture of the NPC, based on the Saccharomyces cerevisiae quaternary structure. Grey and green shapes represent core scaffold Nups and FG-Nups, respectively, identified by DeGrasse et al., 2009 [27]. White shapes represent subcomplexes for which components were not identified in that earlier proteomic screen. (C) Direct visualization of the GFP-tagged newly identified Nups confirms that they exhibit the punctate nuclear rim localization characteristic of NPCs. The corresponding 4’, 6-diamino-2-phenylindoledihidrochloride (DAPI) fluorescence was used to image the DNA (k = kinetoplast, n = nucleus). (D) Secondary structure features and fold prediction of the five newly identified Nups. The y-axis indicates the confidence score of the predicted secondary structure element. Models of fold types are shown on the right, together with potential opisthokont orthologs based on the predicted fold types. RRM, RNA recognition motif; TM, trans-membrane domain. Fold models are based on PDB structures: 1XIP (β-propeller of Nup159), 3P3D (RRM of Nup35), 2KA2 (TM), 1AQ5 (coiled coil), and 4MHC (α-solenoid of Nup192). TbNup152 is approximately 153 kDa but has been assigned 152 to prevent confusion with the well-studied human Nup153.
Fig 2
Fig 2. Affinity isolation of TbNPC subcomplexes.
TbNup nomenclature has been shortened to NupX, with subsequent comigrating Nups simply given their identification number that corresponds to their molecular weight, with the exception of Sec13 (i.e, Nup158, 152 instead of TbNup158, TbNup152). (A) Coomassie-stained SDS-PAGE of GFP-tagged members of the inner ring of the TbNPC. Predicted homologs, predicted fold types, and the GFP-tagged Nup are shown above each gel. The affinity handle (blue ‡) and isolated proteins identified by mass spectrometry are shown on the right of each protein gel. The asterisks designate known contaminants and non-NPC/nuclear envelope proteins as indicated in Fig 1. Full lists are available in S1 Fig. Nup225, 181, 96, 62, 53a, and 53b form a distinct complex with each other. Nup62 exists as two proteins of different sizes that probably reflect allelic variation due to expansion or contraction of FG-repeats. Nup65 associates with Nup96 and 225. Nup144 weakly interacts with Nup89, whilst Nup119 associates with multiple nuclear pore subcomplexes. (B) Affinity isolated members of the outer ring of the TbNPC. Most of the Nups associate with each other, with a few minor exceptions. Nup109 associates weakly with the rest of the complex and is lost in most affinity capture conditions. However, it is a bona fide member of the outer ring, as it affinity isolates the corresponding members of the Nup89 complex. The Nup89 complex also interacts with the lamin analog NUP-1 [42], the nuclear basket Nup110 [27,35], and the FG-Nup98. The presence of Sec13 in both the NPC and COPII complex is highlighted by the affinity capture of Nups as well as the abundant Sec31, a vesicle coat protein that forms a heterotetramer with Sec13 [43], when Sec13-GFP is used as the affinity handle. (C) FG-Nup64 and 98 associate with multiple NPC subcomplexes. Nup75 only interacts with Nup64 and 98, suggesting a close association of these three FG-Nups. (D) Affinity isolation of Nup76 and several FG-Nups with their interacting partners. Nup76 associates with FG-Nups 140, 149, and several members of the outer ring complex. Additionally, the mRNA export factor Mex67 associates with this subcomplex.
Fig 3
Fig 3. Summary of affinity capture of all Nups.
Affinity capture data in Fig 2 are summarized in the above figure and delineate the discrete subcomplexes and the connections between them that define TbNPC subcomplexes and higher order architecture. TbNups that form homodimers are noted, as are the putative yeast and human orthologs of each Nup. The peach color on the label represents outer ring Nups. Purple = inner ring α-solenoids and β-α Nups, blue and pink represent the linker Nups, green = FG-Nups, yellow = nuclear basket Nups, and white = TbNup48/ALADIN, which was not characterized in this study due to our inability to find co-isolating Nups, despite testing several affinity isolation conditions.
Fig 4
Fig 4. Determination of the relative NPC location of each subcomplex.
(A) Immunogold electron localization of GFP-tagged Nups using polyclonal anti-GFP rabbit antibodies to determine relative positions of Nups within the TbNPC (Methods). We picked NPCs sectioned perpendicular to the NE plane, selected a radius of 300 nm around the estimated center of each NPC, and excised each image (S2 Fig). We then aligned and created a superimposed montage of several excised NPC images [6,58]. Graduated lines adjacent to each iEM montage are scaled to represent distances of 50 nm. Major features of each montage are represented in the illustration on the right: NE, nuclear envelope; NPC, nuclear pore complex; N, nucleoplasm; and C, cytoplasm. (B) Statistical analysis of relative locations of select TbNups within the TbNPC, based on the distribution of gold particles from various iEM montages. X and Y positions of gold particles from iEM montages for each selected Nup were measured, from which the Z- and R- (cylindrical rotational axis of the NPC) axes were calculated and displayed in a tabulated form (see S3 Fig, S1 File, and the full table in S2 Table). Z average values are positive or negative to represent localizations above and below the midplane of the NPC. TbNup110 only has a negative value, as it clearly localizes to the nucleoplasm only. Abbreviations: ave (average), Err (error), N(R) (number of gold particles used to calculate the R-axis), N(Z) (number of gold particles used to calculate the Z-axis), NPCs (number of NPCs used to generate either the N(R) or N(Z) for each selected TbNup). (C) Illustrated representation of the relative position of each Nup within the TbNPC. Nup64, 98, and Nup119 are centrally located, whereas Nup62, 76, and 89 appear to be positioned further away from the central channel. The nuclear basket TbNup110 has a clear nucleoplasmic localization. R- and Z-axes errors are plotted based on the 95% level of a peak finding algorithm [6].
Fig 5
Fig 5. Membrane anchoring and the core module of the TbNup89 complex.
(A) TbNup65 is a TM containing protein. (i) Western blot showing sodium carbonate extraction of TM proteins [64], confirming that TbNup65 and Tb927.4.4760—a nuclear envelope and Golgi marker protein—are TM proteins, as they are predominantly recovered in the pellet (Pel) whilst the non-TM α-solenoid TbNup89 is predominantly recovered in the supernatant (Sup). (ii) An illustration of the predicted secondary structure and the differences in nuclear membrane interaction between TbNup65 and its yeast, human, and plant orthologs (ScNup53, HsNup35, and AtNup35, respectively). The opisthokont and plant Nup53/35 are mainly disordered (Disopred), unlike the trypanosome Nup65 that has several structured regions. (B) The Nup89 complex is comprised of eight proteins (including TbNup109) that can be further reduced into a core module consisting of just four proteins when the stringency of the extraction buffer is increased. A schematic of the outer ring as well as subcomplexes is shown. Nup41 and Sec13 are beta propellers, Nup82 and 89 are alpha solenoids, Nup109, 132 and 152 are beta/alphas and Nup158 is a FG-Nup/alpha solenoid.
Fig 6
Fig 6. The interactions of an evolutionarily conserved mRNA exporter with the TbNPC and Ran.
(A) TbMex67 interacts primarily with the Nup76 complex and several components of the TbNup89 complex. TbMex67 also interacts with Ran, Ran Binding Protein 1 (RanBP1), and a GTPase activating protein (TbTBC-RootA) [104] that shows similarity to Rab GTPase Activating Proteins (GAPs) by protein domain prediction. These interactions are suggestive of a role for the Ran gradient in the export of bulk polyA mRNA export. Under low stringency conditions, the interaction between TbMex67 and the TbNPC is clearly observed in a manner reminiscent to that of yeast Mex67 [40]. (B) Models of the trypanosome NTF2, TbMtr2, and the NTF2 domain of TbMex67 were generated using I-TASSER, resulting in C-scores of 1.14, -0.96, and 0.95, respectively. The C-score is used to assess the quality of a model generated by I-TASSER [105]. Its calculation is based on the Z-score of individual threading alignments and the convergence parameters of the I-TASSER assembly simulations. C-scores range between -5 and 2; the closer the score to 2, the higher the confidence in the model generated. The C-scores generated for our models are closer to 2, reflecting high confidence in the models generated. TbNTF2 is capable of binding Ran, based on an accessible potential Ran-binding pocket [106,107], whereas the potential Ran-binding pocket in TbMtr2 and the NTF2 domain of TbMex67 are predicted to be inaccessible, based on structural modeling using I-TASSER. Significantly, this mirrors the situation in yeast and vertebrates, suggesting that Ran binding may not be direct and probably requires the other Ran interacting proteins such as RanBP1 and TbTBC-RootA.
Fig 7
Fig 7. Model of the TbNPC and a putative role of Ran in mRNA export.
(A) A model of the TbNPC compared to the yeast NPC. Only one copy of the inner ring is illustrated for simplicity. The anchoring mechanism of the TbNPC is provided by a single inner ring Nup (TbNup65) that in yeast (ScNup53/59) interacts with the NE via an ALPS motif. Trypanosomes lack the whole pore membrane ring comprised of Pom152 (GP210 in humans and plants), Pom34, and NDC1 [5,6]. The TbNPC is largely symmetric, with asymmetry provided by its nucleoplasmic interactions through two nuclear basket Nups that are half the size of their opisthokont analogs [35]. Significantly, there are no clear orthologs of Dbp5 and Gle1, coincident with the lack of cytoplasmic or nucleoplasmic biased FG-Nups in trypanosomes. Instead, TbNup76, the candidate ortholog of the cytoplasm-specific Nup82/88 in opisthokonts, localizes to both faces of the NPC. (B) Left, model highlighting the conserved inner ring core (blue) and differences in asymmetry (red) in excavates and opisthokonts as represented by trypanosomes and yeast. Orthologs of cytoplasmic Nups or mRNA remodeling factors are absent from trypanosomes. Right, affinity capture of the conserved nonkaryopherin RNA exporter Mex67 co-isolates Ran, suggesting a putative role for the GTPase Ran in mRNA export in trypanosomes (see Fig 6A). Bulk polyA mRNA export in opisthokonts is driven by ATP through the actions of the ATP-dependent DEAD box helicase DBP5, RNA export factor Gle1, and inositol hexakisphosphate (IP6) [22].

Comment in

Similar articles

See all similar articles

Cited by 35 articles

See all "Cited by" articles


    1. Szathmary E, Smith JM. The major evolutionary transitions. Nature. 1995;374(6519):227–32. 10.1038/374227a0 . - DOI - PubMed
    1. Kabachinski G, Schwartz TU. The nuclear pore complex—structure and function at a glance. J Cell Sci. 2015;128(3):423–9. - PMC - PubMed
    1. Alber F, Dokudovskaya S, Veenhoff LM, Zhang W, Kipper J, Devos D, et al. Determining the architectures of macromolecular assemblies. Nature. 2007;450(7170):683–94. Epub 2007/11/30. [pii] 10.1038/nature06404 . - DOI - PubMed
    1. Alber F, Dokudovskaya S, Veenhoff LM, Zhang W, Kipper J, Devos D, et al. The molecular architecture of the nuclear pore complex. Nature. 2007;450(7170):695–701. Epub 2007/11/30. [pii] 10.1038/nature06405 . - DOI - PubMed
    1. Cronshaw JM, Krutchinsky AN, Zhang W, Chait BT, Matunis MJ. Proteomic analysis of the mammalian nuclear pore complex. J Cell Biol. 2002;158(5):915–27. Epub 2002/08/28. 10.1083/jcb.200206106 - DOI - PMC - PubMed

Publication types

MeSH terms


LinkOut - more resources