The Oncogenic Fusion Proteins SET-Nup214 and Sequestosome-1 (SQSTM1)-Nup214 Form Dynamic Nuclear Bodies and Differentially Affect Nuclear Protein and Poly(A)⁺ RNA Export

Abstract

Genetic rearrangements are a hallmark of several forms of leukemia and can lead to oncogenic fusion proteins. One example of an affected chromosomal region is the gene coding for Nup214, a nucleoporin that localizes to the cytoplasmic side of the nuclear pore complex (NPC). We investigated two such fusion proteins, SET-Nup214 and SQSTM1 (sequestosome)-Nup214, both containing C-terminal portions of Nup214. SET-Nup214 nuclear bodies containing the nuclear export receptor CRM1 were observed in the leukemia cell lines LOUCY and MEGAL. Overexpression of SET-Nup214 in HeLa cells leads to the formation of similar nuclear bodies that recruit CRM1, export cargo proteins, and certain nucleoporins and concomitantly affect nuclear protein and poly(A)⁺ RNA export. SQSTM1-Nup214, although mostly cytoplasmic, also forms nuclear bodies and inhibits nuclear protein but not poly(A)⁺ RNA export. The interaction of the fusion proteins with CRM1 is RanGTP-dependent, as shown in co-immunoprecipitation experiments and binding assays. Further analysis revealed that the Nup214 parts mediate the inhibition of nuclear export, whereas the SET or SQSTM1 part determines the localization of the fusion protein and therefore the extent of the effect. SET-Nup214 nuclear bodies are highly mobile structures, which are in equilibrium with the nucleoplasm in interphase and disassemble during mitosis or upon treatment of cells with the CRM1-inhibitor leptomycin B. Strikingly, we found that nucleoporins can be released from nuclear bodies and reintegrated into existing NPC. Our results point to nuclear bodies as a means of preventing the formation of potentially insoluble and harmful protein aggregates that also may serve as storage compartments for nuclear transport factors.

Introduction

The formation of membraneless organelles by phase separation has been appreciated as an important mechanism for the biogenesis of cytoplasmic or nuclear structures as diverse as nucleoli, P-bodies, stress granules, Cajal bodies, and others (reviewed in Ref. 1). In several neurodegenerative diseases including Alzheimer's, Parkinson's, and Huntington's disease as well as amyotrophic lateral sclerosis and frontotemporal dementia, aberrant proteins form amyloid-like aggregates. In many cases, intrinsically disordered regions of aberrant proteins seem to contribute to phase separation and aggregate formation. Interestingly, such aggregates have been shown to interfere with various nucleocytoplasmic transport pathways (2, 3). Similar observations were made in certain leukemias, where aberrant fusion proteins resulting from gene rearrangements form aggregates, mostly in the nucleus of affected cells. Prominent examples are fusions of nucleoporins, the proteins of the nuclear pore complex (NPC).² As for the neurodegenerative diseases described above, such leukemic fusion proteins inhibit nucleocytoplasmic transport when overexpressed in cells (4).

The NPC is the gate for all trafficking events between the nucleus and the cytoplasm. It consists of ∼30 different nucleoporins, occurring in copy numbers of multiples of eight, building a complex of ∼110 MDa in vertebrate cells (for recent reviews see Refs. 5–7 and references therein). A subset of these nucleoporins is characterized by phenylalanine-glycine (FG) repeats, regions that tend to be intrinsically disordered and interact with nucleocytoplasmic transport receptors (8–10). A prominent FG nucleoporin that forms a complex with Nup88 on the cytoplasmic side of the NPC is Nup214, also known as CAN (11). Nup214 plays a role in the export of a subset of nucleocytoplasmic shuttling proteins (12, 13) and tightly binds the nuclear export receptor CRM1 in an FG repeat-dependent manner (14). CRM1 is the major transport receptor for nuclear export of proteins and RNA-protein complexes (15). In the nucleus, it interacts with cargo proteins carrying a characteristic nuclear export sequence (NES) and with the small GTP-binding protein Ran in its GTP-bound form. This trimeric complex can then translocate through the central channel of the NPC and interact with Nup214 at a terminal binding site on the cytoplasmic side of the nuclear pore (16). GTP hydrolysis on Ran, promoted by the cytoplasmic protein RanGAP, eventually leads to dissociation of the export complex.

Several fusions of Nup214 have been implicated in cases of acute leukemia (17), three of which involve the FG region at the C terminus of the nucleoporin. Breaks on chromosomes 6 and 9 that occur in some forms of acute myeloid leukemia (AML) result in a fusion gene, DEK-Nup214, coding for a 165 kDa chimeric protein (18, 19). DEK is a proto-oncogene involved in chromatin organization and transcription regulation (20, 21). The same breakpoint on chromosome 9, targeting an intron of the Nup214 gene, was also detected in the SET-Nup214 fusion gene (18). This gene codes for a fusion protein of 155 kDa with the identical C-terminal Nup214 fragment as DEK-Nup214 and was found in various cases of AML and T-cell acute lymphoblastic leukemia (T-ALL) (22, 23). The SET gene itself, located on chromosome 9 as well, codes for a nuclear proto-oncogene with multiple functions. SET and DEK might act as opposing factors to regulate chromatin modification (21). Both proteins, DEK-Nup214 and SET-Nup214, form nuclear bodies in overexpressing cells, reminiscent of membraneless organelles like Cajal bodies (24), and interact with CRM1, as shown in co-immunoprecipitation experiments (11). As a result, the SET-Nup214 fusion protein inhibits CRM1-dependent nuclear export of an artificial GFP reporter linked to a NES, an effect that is reversed by overexpressing the export receptor (4). The third protein, found as a fusion with a C-terminal, FG-containing region of Nup214 is sequestosome-1, also known as SQSTM1 or p62 (25). The fusion, which involves chromosome 5, where the SQSTM1 gene is located, and the Nup214 locus on chromosome 9, was identified in a T-ALL patient (25). SQSTM1 has been described as an adapter protein with multiple cellular functions, e.g. in selective autophagy and oxidative stress signaling (26, 27) (reviewed in Ref. 28) and can be found commonly in protein aggregates of polyubiquitinated proteins and autophagosomal structures (29). It also has been described as a nucleocytoplasmic shuttling protein (30) that accumulates in PML bodies upon inhibition of CRM1-dependent export (27). The cell biological consequences of SQSTM1-Nup214 overexpression have not been analyzed so far. We compared here the subcellular localization of SET-Nup214 and SQSTM1-Nup214, their biochemical properties with respect to CRM1 binding, and their effects on nuclear protein and poly(A)⁺ RNA export. We show that SET-Nup214, as well as SQSTM1-Nup214, forms highly dynamic nuclear bodies that also recruit CRM1, export cargo proteins, and certain nucleoporins. The effects of the fusion proteins on poly(A)⁺ RNA and protein export depend on their subcellular localization and are mediated by the fused Nup214 fragments.

Results

SET-Nup214 and SQSTM1-Nup214 Form Distinct Nuclear Bodies

Through chromosomal translocation, the SET-Nup214 protein occurs as a fusion of the N-terminal 270 of 277 amino acid residues of SET and a C-terminal, FG repeat-containing fragment of Nup214 comprising amino acid residues 813–2090. LOUCY, a T-ALL cell line, and MEGAL, a line derived from an AML patient, both contain the SET-Nup214 translocation (22). In both cell lines, SET-Nup214 forms distinct bodies in the nucleus (Fig. 1A). Such bodies could be detected in 70% of the LOUCY cells and 10% of the MEGAL cells, which coincides with a higher expression level of SET-Nup214 in LOUCY compared with MEGAL cells (data not shown). In contrast, no accumulation of SET-Nup214 was seen in K562 cells, an unrelated leukemia cell line (Fig. 1A). Compared with SET-Nup214, SQSTM1-Nup214 contains a shorter C-terminal FG repeat-containing Nup214 fragment comprising residues 1969–2090, fused to the N-terminal 252 of 440 total residues of SQSTM1 (Fig. 1B).

FIGURE 1.

Cellular localization of the oncogenic fusion proteins SET-Nup214 and SQSTM1-Nup214. A, LOUCY, MEGAL, and K562 cells were stained with an antibody against Nup214 and analyzed by confocal microscopy. DNA was stained with DAPI (blue). Size bars, 5 μm. Arrowheads point to nuclear bodies. B, schematic drawing of endogenous proteins, fusion proteins, and Nup214 fragments used in this study. SET (green), SQSTM1 (orange), Nup214 (gray), and relevant amino acid residues are indicated. C and D, HeLa cells were transfected with plasmids coding for N-terminally mCherry-tagged SET, SET-Nup214, SQSTM1, and SQSTM1-Nup214, and the subcellular localization of the tagged proteins was analyzed by immunoblotting using an α-RFP antibody (C) or confocal microscopy (D). The unspecific band below 70 kDa serves as a loading control. Note that FG-containing proteins typically run at a somewhat higher molecular weight than calculated (C). Neither the type nor the position of the tag had an effect on the localization of the fusion proteins (data not shown). E and F, HeLa cells transfected with either Myc-tagged (red) SET-Nup214 (E) or SQSTM1-Nup214 (F) were stained for common marker proteins for cellular compartments (endogenous coilin, PML, LAMP2, LC3, and SQSTM1 as indicated (green)). Plasmids coding for the FusionRed-ER marker and SQSTM1-Nup214-myc were co-transfected; the endoplasmic reticulum (ER) is colored in green for better visualization. D–F, DNA was stained with Hoechst (blue). Size bars, 20 μm. Arrowheads point to nuclear bodies. The enlarged areas (insets) correspond to a 7-fold magnification of the indicated regions.

We overexpressed mCherry-tagged fusion proteins and the original full-length proteins, SET and SQSTM1, in HeLa cells to compare their subcellular localization pattern. To verify that full-length proteins were produced, lysates of cells transfected with the corresponding N-terminally tagged mCherry constructs were analyzed by SDS-PAGE followed by immunoblotting (Fig. 1C). In accordance with its role as a transcription factor and chromatin regulator, overexpressed SET was evenly distributed throughout the nucleus (Fig. 1D). SET-Nup214, on the other hand, localized in distinct nuclear bodies of variable size and number (Fig. 1D) as described previously (4) and comparable to its localization in leukemia cell lines (Fig. 1A). Endogenous SQSTM1 shuttles between the nucleus and the cytoplasm, as it accumulated in the nucleus in the presence of leptomycin B (LMB) (30, 31), a fungal drug that covalently modifies CRM1 (32) and prevents the formation of nuclear export complexes (33, 34). Overexpressed SQSTM1 localized mainly in cytoplasmic puncta, similar to overexpressed SQSTM1-Nup214 (Fig. 1D). In addition, a small fraction of the SQSTM1-Nup214 could also be observed in nuclear bodies (Fig. 1D). The localization of SQSTM1-Nup214 and its distribution between nuclear and cytoplasmic structures varied between different cells, seeming to depend, among other factors, on the expression level.

Neither SET-Nup214 nor SQSTM1-Nup214 co-localized with any of the common markers of cellular bodies or structures that we tested (Fig. 1, E and F). Nevertheless, both proteins were frequently observed in close vicinity to coilin, a marker for Cajal bodies, and, in the case of SET-Nup214, also to nuclear PML bodies (Fig. 1E). Interestingly, SQSTM1-Nup214 co-localized in some but not all cells with endogenous SQSTM1 in structures resembling autophagosomes, as detected with an antibody that recognizes an SQSTM1 epitope missing in the fusion protein (Fig. 1F). Taken together, the fusion of SET or SQSTM1 to Nup214 leads to an altered localization compared with the individual proteins and to the formation of nuclear bodies that seem to be distinct from classic nuclear Cajal or PML bodies.

Nuclear Bodies Containing SET-Nup214 or SQSTM1-Nup214 Accumulate the Nuclear Export Receptor CRM1

It has been shown previously that CRM1 localizes to SET-Nup214-containing nuclear bodies in transfected HeLa cells (4). Indeed, when we overexpressed SET-Nup214 in HeLa cells, CRM1 also accumulated in SET-Nup214-containing nuclear bodies (Fig. 2A). In the nucleocytoplasmic transport, the FG repeat regions of nucleoporins promote the formation of nuclear export complexes containing CRM1, RanGTP, and export cargo (14, 35). This is also shown for the FG repeat region of Nup214 (14). Strikingly, SET-Nup214-containing nuclear bodies were not observed in cells treated with the CRM1 inhibitor LMB, a drug that prevents export complex formation. Instead, SET-Nup214 and CRM1 were evenly distributed throughout the nucleus (Fig. 2A). These results suggest that the nuclear bodies may contain bona fide nuclear export complexes. We therefore also analyzed the subcellular localization of Ran. Even though Ran is part of the LMB-sensitive CRM1 export complexes, we did not observe an accumulation of Ran in SET-Nup214- and CRM1-containing nuclear bodies (Fig. 2A).

FIGURE 2.

Influence of SET-Nup214 and SQSTM1-Nup214 on the localization of the nuclear export receptor CRM1. A and B, HeLa cells were transfected with plasmids coding for Myc-tagged (red in merged pictures) SET-Nup214 (A) or SQSTM1-Nup214 (B) and stained with antibodies against CRM1 or Ran (green in merged pictures). Size bars, 20 μm. C, LOUCY cells were stained with antibodies against Nup214 (red) and CRM1 (green). Size bars, 5 μm. A–C, cells were incubated with (+) or without (−) 10 nm (HeLa cells) or 20 nm (LOUCY cells) LMB for 3 h, as indicated, and analyzed by confocal microscopy. DNA was stained with DAPI or Hoechst (blue). Arrowheads exemplify nuclear bodies with co-localized Nup214 fusion protein and CRM1.

The major portion of SQSTM1-Nup214 localized to cytoplasmic puncta, although the protein could also be observed in nuclear bodies (Fig. 2B). Similar to SET-Nup214 bodies, the nuclear SQSTM1-Nup214 bodies also accumulated CRM1 and were dissolved upon treatment of cells with LMB (Fig. 2B). In contrast to nuclear SQSTM1-Nup214 bodies, cytoplasmic puncta of SQSTM1-Nup214 were not affected by LMB treatment. Like SET-Nup214, SQSTM1-Nup214 did not affect the localization of Ran (Fig. 2B).

We next tested whether SET-Nup214 nuclear bodies in leukemia cells lines also accumulated CRM1 and were sensitive to LMB treatment. Indeed, the nuclear bodies in LOUCY cells also co-localized with CRM1 and dissolved upon treatment with LMB (Fig. 2C). These results suggest that the SET-Nup214 nuclear bodies in overexpressing HeLa cells and in patient-derived leukemia cells are very similar in nature.

SET-Nup214 and SQSTM1-Nup214 Differentially Inhibit Nuclear Protein and poly(A)⁺ RNA Export

We next investigated whether nuclear export of CRM1-dependent cargo proteins would be disrupted in cells expressing SET-Nup214 or SQSTM1-Nup214 using GFP-tagged versions of the established CRM1 cargoes HIV-1 Rev and snurportin 1 (SPN1). Both cargo proteins localized exclusively to the cytoplasm in control cells and accumulated in the nucleus upon treatment of the cells with LMB (compare Fig. 3, A and B). Nuclear export of both proteins was disrupted in cells containing SET-Nup214 or SQSTM1-Nup214 nuclear bodies (Fig. 3, A and B), leading to nuclear accumulation of the cargo proteins. In addition, the accumulation of HIV-1 Rev and SPN1 in the nuclear bodies themselves was also observed in some cells. It was noteworthy that the localization of SPN1 was not affected in all of the SQSTM1-Nup214-positive cells, and the extent of the effect seemed to be related to the expression level of the fusion protein (Fig. 3B; see also Fig. 6, G and H). Upon LMB treatment and dissolution of the nuclear bodies, the localization of the cargo proteins in cells with or without SET-Nup214 was indistinguishable (Fig. 3, A and B). SET-Nup214 nuclear bodies also disrupted the export of endogenous RanBP1, an established CRM1 cargo (36) (Fig. 3, A and C), whereas SQSTM1-Nup214 did not affect RanBP1 (Fig. 3B).

FIGURE 3.

Inhibition of nuclear export by SET-Nup214 and SQSTM1-Nup214. A and B, HeLa cells were co-transfected with plasmids coding for Myc-tagged (red) SET-Nup214 (A) or SQSTM1-Nup214 (B) and GFP-tagged reporter proteins HIV-1 Rev and SPN1 (green), respectively. For analysis of endogenous RanBP1 (green), cells were subjected to indirect immunofluorescence. The influence of SET-Nup214 and SQSTM1-Nup214 (red) on poly(A)⁺ RNA localization (green) was analyzed by in situ hybridization. DNA was stained with Hoechst (blue). Size bars, 20 μm. C, quantification of the cellular localization of RanBP1 in 31 SET-Nup214-expressing cells compared with 55 untransfected cells. The blue dots represent the individual data points. Error bars indicate mean ± S.D. The gray area indicates a predominant cytoplasmic localization (nuclear/cytoplasmic signal <1). D, LOUCY cells were stained with antibodies against Nup214 (red) and RanBP1 (green). DNA was stained with DAPI (blue). Size bar, 5 μm. A–D, cells were analyzed by confocal microscopy. Where indicated, they were treated with 10 nm LMB for 1 h (HeLa) or 20 nm LMB for 3 h (LOUCY). Arrowheads point to nuclei with nuclear bodies and accumulated cargo.

FIGURE 6.

Effects on nuclear export are mediated by the Nup214 part and depend on the localization of the fusion protein. A, schematic drawing of the two fusion constructs with swapped Nup214 fragments, SET-Nup214(1969–2090) and SQSTM1-Nup214(813–2090). B–F, HeLa cells were transfected with plasmids coding for Myc-tagged SET-Nup214(1969–2090) or SQSTM1-Nup214(813–2090) and analyzed by immunoblotting using an α-Myc antibody (B) or confocal microscopy (C–F). C–F, the fusion proteins were detected by indirect immunofluorescence using an anti-Myc antibody (red). C, cell were stained with an antibody against endogenous CRM1. D and E, cells were co-transfected with the cargo proteins dGFP-HIV-1-Rev (D) and GFP-SPN1 (E). F, the localization of poly(A)⁺ RNA was analyzed by in situ hybridization. C–F, size bar, 20 μm. G and H, quantification of the results in D and E, respectively. The bars show the mean distribution of 100 cells/sample over three independent experiments. Error bars, standard deviation. (*), ∼90% of the cells in either category show an accumulation of dGFP-HIV-1 Rev in the nucleoli; (**), ∼20% of the cells with a predominant cytoplasmic signal also show an accumulation of dGFP-HIV-1 Rev in the nucleoli; (***), ∼half of the cells with a predominant cytoplasmic localization of GFP-SPN1 also exhibit co-localization of the reporter protein with SQSTM1-Nup214 in nuclear bodies; (****), GFP-SPN1 co-localizes with the cytoplasmic SQSTM1-Nup214(813–2090) bodies. N > C, mainly nuclear; N = C, equally distributed between the nucleus and the cytoplasm; N < C, mainly cytoplasmic.

Overexpression of Nup214 has been shown previously to lead to nuclear accumulation of mRNA (37). We therefore investigated the effect of SET-Nup214 and SQSTM1-Nup214 nuclear bodies on poly(A)⁺ RNA localization by fluorescence in situ hybridization (Fig. 3, A and B). In cells expressing SET-Nup214, poly(A)⁺ RNA accumulated in the nucleus forming aggregates that did not overlap with SET-Nup214 nuclear bodies. Upon LMB-induced disassembly of SET-Nup214 nuclear bodies, nuclear poly(A)⁺ RNA aggregates were still visible (data not shown). LMB itself had no effect on poly(A)⁺ RNA localization (data not shown), in agreement with previous observations (33, 38, 39). These results suggest that SET-Nup214 inhibits nuclear poly(A)⁺ RNA export in a CRM1-independent manner. In SQSTM1-Nup214-expressing cells, by contrast, the poly(A)⁺ RNA localization pattern was unchanged (Fig. 3B).

We next verified the physiological relevance of the export inhibition by SET-Nup214 through analysis of the RanBP1 localization in LOUCY cells (Fig. 3D). Endogenous RanBP1 clearly co-localized with the SET-Nup214 in nuclear bodies that were dissolved upon the addition of LMB.

Binding of SET-Nup214 and SQSTM1-Nup214 to the Export Receptor CRM1

The accumulation of the nuclear export receptor CRM1 in SET-Nup214-as well as SQSTM1-Nup214-containing nuclear bodies raised the questions of whether CRM1 binds directly to the fusion proteins and whether their binding mechanisms are similar. First, we performed co-immunoprecipitation experiments from lysates of cells expressing Myc-tagged fusion proteins. Endogenous CRM1 was co-immunoprecipitated with Myc-SET-Nup214 only in the presence of GTP-loaded RanQ69L, a Ran mutant that is resistant to RanGAP-induced GTP hydrolysis (40) (Fig. 4A), suggesting preferential binding of the Nup214 fusion protein to RanGTP-containing export complexes. This result is in agreement with previous findings, where co-immunoprecipitation of endogenous Nup214 with CRM1 was promoted by RanQ69L (16). Under the same conditions, CRM1 could not be co-immunoprecipitated with SQSTM1-Nup214 (Fig. 4B), suggesting a lower affinity of these two proteins. We therefore used purified components to analyze the interaction of SQSTM1-Nup214 with CRM1. First, we performed binding assays with immobilized GST-CRM1 and MBP-tagged SQSTM1-Nup214. The binding of CRM1 to the FG regions of Nup214 was enhanced in the presence of RanGTP and an export cargo (14). As shown in Fig. 4C, RanQ69L-GTP and an NES peptide acting as an artificial cargo promoted the binding of the fusion protein to CRM1 when added together to the reaction. The addition of MBP-tagged SQSTM1 did not lead to reduced binding of MBP-SQSTM1-Nup214, whereas a short Nup214 fragment (amino acids 1916–2033) efficiently competed with MBP-SQSTM1-Nup214 for binding to immobilized CRM1. In a similar set of experiments, we analyzed the binding of MBP-SQSTM1-Nup214 and CRM1 to immobilized RanGDP or RanGTP (Fig. 4D). Again, complex formation was RanGTP-dependent, promoted by an NES peptide, and reduced by a competing Nup214 fragment. These results suggest that the binding of SQSTM1-Nup214 to CRM1 is dominated by the nucleoporin part of the fusion protein and that CRM1-binding to both SET-Nup214 and SQSTM1-Nup214 depends on RanGTP. Because we used proteins expressed in bacteria in these experiments, we can of course not exclude the possibility that protein-protein interactions are further affected by post-translational modifications in vivo.

FIGURE 4.

Binding of SET-Nup214 and SQSTM1-Nup214 to the export receptor CRM1. A and B, co-immunoprecipitation (IP) of CRM1 with Myc-tagged SET-Nup214 (A) or SQSTM1-Nup214 (B). HeLa cells were transfected with plasmids coding for Myc-SET-Nup214 or Myc-SQSTM1-Nup214. After 24 h, the cells were lysed and subjected to co-immunoprecipitations with either mouse IgG or anti-Myc antibody in the absence or presence of RanQ69L-GTP. The precipitated proteins were analyzed by SDS-PAGE and immunoblotting. Note that no input signal was observed for Myc-tagged SET-Nup214 or SQSTM1-Nup214 due to low expression levels. GST-CRM1 (C) or GST-Ran (D) loaded with either GDP or GTP was immobilized on beads and incubated with an excess of MBP-SQSTM1-Nup214, CRM1, RanQ69L-GTP, MBP-SQSTM1, Nup214(1916–2033), or NES peptide as indicated. The bound proteins were analyzed by SDS-PAGE followed by Coomassie staining. C, lower and upper parts of the gel are shown at different intensities and background settings to improve visualization of the Nup214 fragment.

The Effects of the Nup214 Fusion Proteins Depend on Their Nup214 Parts and Their Subcellular Localization

The binding assays described above suggest that the interaction of the fusion proteins and CRM1 is mediated by the Nup214 part. We have shown previously that long fragments derived from the FG-containing part of Nup214 tend to have stronger inhibitory effects on nuclear export compared with shorter ones (14). Hence, the longer FG-containing portion of SET-Nup214 could explain the stronger binding to CRM1 (Fig. 4) and the more prominent effect on protein and poly(A)⁺ RNA export compared with SQSTM1-Nup214 (Fig. 3). We therefore analyzed the influence of the individual parts of SET-Nup214 and SQSTM1-Nup214 in further detail. In co-transfection experiments with the CRM1 cargoes HIV-1 Rev (Fig. 5, A–C) and SPN1 (Fig. 5, D–F), we investigated the full-length proteins SET and SQSTM1 as well as the Nup214 regions present in the fusion proteins and containing amino acid residues 813–2090 and 1969–2090, respectively. N-terminally mCherry-tagged versions of SET and SQSTM1 did not affect the export of HIV-1 Rev (Fig. 5A) or SPN1 (Fig. 5D). The Nup214 fragments were expressed as RFP-tagged fusion proteins with a C-terminal classical nuclear localization signal (cNLS) to ensure accumulation in the nucleus. Both Nup214 fragments inhibited nuclear export of HIV-1 Rev (Fig. 5, B and C) and SPN1 (Fig. 5, E and F), resulting in accumulation of the export cargo proteins in the nucleus. In some cells, Nup214(813–2090) localized to nuclear bodies that resembled the SET-Nup214 bodies. Interestingly, the effect of Nup214(813–2090) on the export of SPN1 varied depending on its localization (Fig. 5, E and F). In cells that showed an even distribution of Nup214(813–2090) throughout the nucleus, the export of SPN1 was disrupted and SPN1 localized exclusively to the nucleus. When the Nup214 fragment accumulated in nuclear bodies, nuclear export of SPN1 was not affected and the reporter protein localized predominantly to the cytoplasm (Fig. 5, E and F). This is in contrast to the effect of Nup214(813–2090) on HIV-1 Rev export, which was disrupted even though Nup214(813–2090) accumulated in nuclear bodies (Fig. 5, B and C).

FIGURE 5.

Effects on nuclear export of the individual components of the fusion proteins. HeLa cells were co-transfected with plasmids coding for dGFP-HIV-1 Rev (A–C) or GFP-SPN1 (D–F) and N-terminally mCherry-tagged SET (A and D) or SQSTM1 (A and D) or the Nup214 fragments fused to RFP-cNLS (B and E) as indicated. Cells were analyzed by confocal microscopy. DNA was stained with Hoechst (blue). Size bars, 20 μm. Arrowheads point to nuclear bodies. C and F, quantification of the results in B and E. The bars show the mean distribution of 100 cells/sample over two independent experiments. Error bars, standard deviation. N > C, mainly nuclear; N = C, equally distributed between the nucleus and the cytoplasm; N < C, mainly cytoplasmic.

For a more detailed analysis of the contribution of the individual fusion partners to the observed effects, we designed expression constructs for novel fusion proteins with the Nup214 parts swapped. SET-Nup214(1969–2090) comprises the N-terminal SET fragment (amino acids 1–270) and the short C-terminal Nup214 fragment originally found in SQSTM1-Nup214. SQSTM1-Nup214(813–2090), on the other hand, contains the N-terminal SQSTM1 fragment linked to a long Nup214 fragment as originally found in SET-Nup214 (Fig. 6A). First, HeLa cells were transfected to express the novel fusion proteins, which had the expected size as analyzed by SDS-PAGE (Fig. 6B). Next, we compared the effects of the different fusion proteins on endogenous CRM1. As shown in Fig. 6C, SET-Nup214(1969–2090) was found exclusively in the nuclei of transfected cells. In contrast to the original Set-Nup214, however, nuclear bodies were not observed. Accordingly, CRM1 did not localize to nuclear bodies either, although the export receptor was somewhat concentrated in the nuclei compared with non-transfected cells. SQSTM1-Nup214(813–2090), on the other hand, was largely excluded from the nucleus and no nuclear bodies could be observed. Endogenous CRM1 co-localized to some extent with SQSTM1-Nup214(813–2090) in cytoplasmic dots (Fig. 6C). Next, we analyzed the fusion proteins with respect to the nuclear export of our reporter proteins, GFP-HIV-1 Rev and GFP-SPN1 (Fig. 6, D and E). In agreement with its effect on endogenous CRM1, SET-Nup214(1969–2090) had a clearly inhibitory effect on the nuclear export of both reporter proteins (Fig. 6, D, E, G, and H). Here, the extent of inhibition was similar to that observed for the original SET-Nup214 protein (Fig. 6, G and H). SQSTM1-Nup214(813–2090), by contrast, was clearly less efficient in inhibiting nuclear protein export compared with the original SQSTM1-Nup214 fusion protein, in agreement with its exclusion from the nucleus (Fig. 6, G and H). Of note, some cells expressing SQSTM1-Nup214(813–2090) showed an accumulation of HIV-1 Rev in the nucleoli (Fig. 6D, arrowheads). Finally, we characterized the effects of the fusion protein on the subcellular localization of poly(A)⁺-RNA (Fig. 6F). SET-Nup214(1969–2090) resulted in an accumulation of the RNA in the nuclei of transfected cells similar to that observed previously for SET-Nup214 (Fig. 3A). SQSTM1-Nup214(813–2090) had no effect on poly(A)⁺-RNA distribution. Our combined results (Figs. 3–6) suggest that the Nup214 FG repeats mediate the effect on nuclear export and that the length of the Nup214 part may affect the degree of inhibition of the analyzed export pathways. The ability to exert such effects, however, depends on their N-terminal fusion partner and the resulting subcellular localization of the fusion protein.

SET-Nup214, but Not SQSTM1-Nup214, Affects the Localization of Endogenous Nucleoporins

As the effect of fusion proteins seemed to be, at least in part, mediated by the FG repeat-containing Nup214 portions of the fusion protein, we investigated whether endogenous nucleoporins were affected by the expression of SET-Nup214 and SQSTM1-Nup214 (Fig. 7). Nup88 is localized to the cytoplasmic side of the nuclear pore complex and forms a subcomplex with Nup214 (41, 42). In cells expressing SET-Nup214, Nup88 at the nuclear rim was clearly reduced with a concomitant appearance of the nucleoporin in SET-Nup214 nuclear bodies. The treatment of cells with LMB resulted in a dispersed nuclear localization of the Nup214 fusion protein, as described above, and also a clearly visible Nup88 rim (Fig. 7A). Similar observations were made for Nup62, at least in some cells (Fig. 7A). Interestingly, Nup62 was dispersed mostly in the nucleoplasm in LMB-treated cells expressing SET-Nup214. Endogenous Nup98, shown previously to accumulate in nuclear bodies under certain conditions (43), was not recruited to SET-Nup214 bodies. In experiments where Nup98 was overexpressed together with SET-Nup214, however, both proteins co-localized in nuclear bodies (data not shown). Hence, we cannot exclude the possibility that endogenous Nup98 is present as well but not detected under our experimental conditions. Neither Nup358, a nucleoporin at the cytoplasmic side of the NPC (44, 45), nor Tpr, a component of the nuclear basket (46), was affected by SET-Nup214 expression. SQSTM1-Nup214 did not influence the localization of any of the endogenous nucleoporins tested (Fig. 7B). Together, a selected set of nucleoporins can be recruited to SET-Nup214 nuclear bodies, depleting them from the nuclear pore.

FIGURE 7.

SET-Nup214, but not SQSTM1-Nup214, influences the localization of endogenous nucleoporins. Myc-tagged SET-Nup214 (A) and SQSTM1-Nup214 (B) were expressed in HeLa cells, and the localization of endogenous nucleoporins was analyzed by confocal microscopy. The fusion proteins were detected by indirect immunofluorescence using an anti-Myc antibody (red). Cells were treated with 10 nm LMB for 3 h, as indicated. The enlarged areas (insets) correspond to a 7-fold magnification of the indicated regions. Size bars, 20 μm.

Nuclear SET-Nup214 Bodies Are Highly Dynamic

We showed that SET-Nup214- and SQSTM1-Nup214-containing nuclear bodies dissolve upon treatment of cells with LMB (Fig. 2). This, along with the observation that a subset of FG nucleoporins is recruited specifically to SET-Nup214 but not to SQSTM1-Nup214 nuclear bodies (Fig. 7), indicates that nuclear bodies do not represent aggregates of insoluble protein but rather accumulations of various functional proteins. We therefore investigated the dynamics of nuclear bodies in further detail. First, we analyzed whether SET-Nup214 and SQSTM1-Nup214 were soluble in physiological buffers. Cells transfected with either SET-Nup214 or SQSTM1-Nup214 were lysed by repeated freeze-thaw cycles, and soluble and insoluble proteins were separated by centrifugation (Fig. 8A). SET-Nup214, CRM1, and Ran were found only in the soluble fraction, demonstrating that the nuclear bodies do not represent aggregates of denatured proteins. SQSTM1-Nup214, by contrast, was found predominately in the insoluble fraction, suggesting that the cytoplasmic SQSTM1-Nup214 puncta might either be aggregates of denatured, insoluble protein or that the fusion protein might be bound to a membrane. LMB did not lead to any changes in the fractionation behavior of the Nup214 fusion proteins.

FIGURE 8.

Nuclear bodies of SET-Nup214 are mobile and dynamic. A, HeLa cells transfected with plasmids coding for Myc-tagged SET-Nup214 or SQSTM1-Nup214 were treated with 10 nm LMB for 1 h, as indicated, and subjected to subcellular fractionation. Equivalent amounts of pellet (P) or soluble fractions (S) were separated by SDS-PAGE followed by immunoblotting. B, cells were transfected with plasmids coding for mCherry-SET-Nup214 and GFP-histone 2A (GFP-H2A) and treated with thymidine and nocodazole. Typical examples at different stages of the cell cycle are shown. Size bar, 10 μm. C and D, HeLa cells were transfected with a plasmid coding for mCherry-SET-Nup214 and subjected to FRAP analysis. C, representative images of cells used for FRAP analysis. Time points are indicated in brackets. The arrowheads indicate the bleached nuclear body (black), another nuclear body from the bleached cell (dark gray), and a nuclear body from a non-bleached cell (light gray). Size bar, 10 μm. D, quantitative analysis of nuclear body fluorescence during FRAP. Depicted is the mean relative fluorescence of nuclear bodies from 24 individual cells over a time course of 400 s. Error bars indicate mean ± S.D. The gray area indicates the bleaching interval. E, HeLa cells were transfected with a plasmid coding for mCherry-SET-Nup214 and treated with 10 nm LMB and 100 μg/ml cycloheximide (CHX) for 0, 30, or 120 min as indicated. After fixation, the localization of mCherry-SET-Nup214 (red) and endogenous Nup88 (green) was analyzed by confocal microscopy. Arrowheads point to nuclear envelopes with reintegrated Nup88 after dissolving of the nuclear bodies. Size bar, 20 μm.

After this biochemical characterization, we investigated the dynamics of SET-Nup214 nuclear bodies in intact cells, starting with an analysis of their behavior throughout the cell cycle. Cells were treated with thymidine followed by a release period, nocodazole treatment, and another washing step to enrich for cells in different stages of the cell cycle (Fig. 8B). To simplify the analysis, cells were co-transfected with a plasmid coding for GFP-histone 2A together with the plasmid coding for Myc-tagged SET-Nup214. In interphase, SET-Nup214 localized to nuclear bodies as shown throughout this study. GFP-histone 2A was evenly distributed in the nucleus and the nucleoli. Toward the end of G₂ phase, the chromosomes condensed, the nuclear bodies dissolved, and SET-Nup214 was detected at the nuclear rim (data not shown). During mitosis, as indicated by the fully condensed, GFP-histone 2A-positive chromosomes, SET-Nup214 was distributed evenly in the whole cell. After mitosis, the process seemed to be reversed. SET-Nup214 accumulated at the nuclear rim followed by de novo formation of nuclear bodies (Fig. 8B). Very similar results were obtained when we analyzed the cells by indirect immunofluorescence, detecting the mitosis marker phosphohistone 3 (data not shown). Together, these results show that nuclear bodies, as formed in cells expressing SET-Nup214, can dissolve and reform during the cell cycle.

In light of the observed dynamics of the nuclear bodies as a whole, we next analyzed by live cell imaging whether there was any exchange of proteins between individual nuclear bodies and/or the nucleoplasm during interphase. In the time frame of 10 min, the nuclear bodies constantly moved through the nucleus in all directions (data not shown). In FRAP (fluorescence recovery after photobleaching) experiments, we bleached one nuclear body and measured its fluorescence intensity as well as the fluorescence intensity of two other nuclear bodies over a time course of ∼400 s (Fig. 8, C and D). The fluorescence intensity of the bleached nuclear body partially recovered within 400 s. Simultaneously, the fluorescence of a second body within the same cell that had not been bleached decreased over time. The fluorescence of a third body located in a peripheral cell within the same frame was used as a reference. This indicates that SET-Nup214 shuttles between individual nuclear bodies. As the fluorescence of the bleached body recovered by only about 30%, it seems that a significant portion of the reporter protein within the nuclear body remained immobile over the time course of the experiment.

Finally, we tested whether nucleoporins that accumulate in nuclear SET-Nup214 bodies could move back to the NPC. SET-Nup214-expressing cells were treated with LMB to dissolve the nuclear bodies and, simultaneously, with cycloheximide to inhibit protein biosynthesis. The localization of endogenous Nup88 was analyzed at different time points (Fig. 8E). At 30 min, Nup88 clearly co-localized with SET-Nup214 bodies. After 60 min (data not shown) of LMB/cycloheximide treatment, dispersal of the nuclear bodies and an even distribution of Nup88 throughout the nucleus was observed. After 120 min, the majority of Nup88 was found at the nuclear rim similar to neighboring cells that did not express SET-Nup214, indicating that the nucleoporin had been partially reintegrated into the NPC. This result suggests that nucleoporins that are sequestered in SET-Nup214 bodies can return to their original localization and possibly fulfill their function as a component of the NPC.

Discussion

Interaction of CRM1 with Nup214 Fusion Proteins

We have shown previously that the FG-rich regions in Nup214 are crucial for the interaction with the export receptor CRM1 (14, 35). The two oncogenic Nup214 fusion proteins, SET-Nup214 and SQSTM1-Nup214, contain fragments of the C-terminal part of Nup214, which is rich in FG repeats. For SET-Nup214, which contains the entire FG region of Nup214 (amino acids 813–2090 with a total of 42 FG repeats), we expected to find binding characteristics with respect to CRM1 very similar to those of the endogenous nucleoporin. Indeed, the fusion protein bound to the export receptor in a RanGTP-dependent manner (Fig. 4A). Unfortunately, we were unable to express SET-Nup214 in bacteria in a soluble form, precluding a more detailed analysis. SQSTM1-Nup214, by contrast, contains a rather short C-terminal fragment of Nup214 (amino acids 1969–2090) with only 14 FG repeats. A very similar Nup214 fragment (amino acids 1975–2090) inhibited CRM1-dependent nuclear export in a previous study (14), suggesting that this region of Nup214 is sufficient to bind to CRM1. Indeed, we observed comparable RanGTP- and NES-dependent binding of purified SQSTM1-Nup214 to CRM1 (Fig. 4, C and D). Thus, our results show that the oncogenic fusion proteins of Nup214 interact with CRM1 in a manner that is dominated by their nucleoporin part, corroborating the original findings by Saito et al. (4).

Nuclear Bodies Containing Nup214 Fusion Proteins

Both SET-Nup214 and SQSTM1-Nup214 aggregate in nuclear bodies that are distinct from Cajal and PML bodies. However, several other types of nuclear bodies have been described previously that share certain characteristics with SET-Nup214- or SQSTM1-Nup214-containing nuclear bodies. One of the proteins forming similar nuclear bodies is the FG repeat-containing nucleoporin, Nup98. Comparable with Nup214, oncogenic Nup98 fusion proteins have been described that form nuclear dots and also lead to the inhibition of CRM1-dependent nuclear export (47). Furthermore, GLFG bodies containing Nup98 with its characteristic GLFG domain can be observed in certain HeLa cell lines (43, 48). GLFG bodies formed by overexpressed Nup98 also recruit CRM1 and are dissociated upon the addition of LMB (47). CNoB (CRM1 nucleolar bodies) are another type of nuclear body. CNoB are found in the nucleoli of cells overexpressing the transcription factor CPEB1. SET-Nup214 and SQSTM1-Nup214 nuclear bodies are largely absent from the nucleoli, suggesting that they are distinct from CNoB. Like the nuclear bodies formed by the nucleoporin fusion proteins, CNoB also disappear upon treatment with LMB (49). Finally, nuclear bodies containing endogenous SQSTM1 have been documented (27). They are also found in close vicinity to and even partially co-localizing with PML and Cajal bodies. However, contrary to the other types of nuclear bodies described above, their formation is induced by LMB.

The cellular localization of a large number of proteins to different nuclear bodies has been analyzed on a systematic scale by Fong et al. (50). Nuclear transport receptors or nucleoporins, however, were not investigated in that study. We showed here that the nuclear bodies formed by SET-Nup214 or SQSTM1-Nup214 can accumulate various other proteins, e.g. the endogenous FG nucleoporins Nup88 and Nup62. Other nucleoporins, however, were absent from the Nup214 nuclear bodies. In addition to certain FG nucleoporins, the nuclear bodies also recruited the nuclear export receptor CRM1 and CRM1 cargo proteins like SPN1, HIV-1 Rev, and RanBP1. For nuclear export, CRM1 has to cooperatively bind the cargo and GTP-loaded Ran. Interestingly, SET-Nup214- and SQSTM1-Nup214-containing nuclear bodies failed to accumulate Ran above general nuclear levels, suggesting that it is present in the nuclear bodies but does not reach higher concentrations than in the nucleoplasm. Thus, we cannot exclude the possibility that nuclear bodies contain fully assembled export complexes or individual components of such complexes at different stoichiometries. Cargo binding clearly involves the characteristic NES-binding cleft of CRM1, as the nuclear bodies are sensitive to LMB, which covalently modifies an amino acid in this region of the export receptor (32). Although no CRM1-containing nuclear bodies were observed in the absence of the Nup214 fusion proteins, we cannot exclude the possibility that small amounts of CRM1 associate with defined regions of the chromatin, which could serve as seeding points for the formation of nuclear bodies upon overexpression of our fusion proteins. Similar observations were published very recently (51), where CRM1 was found to bind to Hox gene clusters.

Dynamics of Nuclear Bodies and Inhibition of Nuclear Transport

Our results show that the Nup214-containing nuclear bodies are not simply accumulations of superfluous proteins prone to aggregation and, possibly, degradation. Instead, the stored proteins are in a dynamic equilibrium with the nucleoplasm, other nuclear bodies of the same type, and even the NPC (Fig. 8, C–E). What drives the formation of SET-Nup214- or SQSTM1-Nup214-containing nuclear bodies? A simple explanation might be an elevated concentration of molecules prone to form membraneless organelles. At a certain concentration, multidomain proteins with intrinsically disordered regions tend to undergo a phase transition leading to the formation of lipid droplets in the cytoplasm or the nucleus (1). Depending on the characteristics of the initiating molecule (e.g. a Nup214 fusion protein), these droplets are expected to attract interacting proteins like CRM1, its transport cargo proteins, or even other FG nucleoporins. The stored proteins can be released (and reused) again, e.g. when the cellular concentration of the seeding molecules drops again. On a general note, cells have to deal with nucleoporins that tend to be imported into the nucleus upon synthesis as a result of their FG-rich regions, which may facilitate translocation through existing NPC. For Nup214 (or its fusion proteins) with its high affinity for CRM1 (35), interaction with the export receptor in the nucleus may prevent deleterious aggregations at high protein concentrations. Transport receptors have been shown previously to inhibit the formation of potentially toxic, amyloid-like aggregates of FG Nup (52). On the other hand, our observation that the nuclear bodies are dissolved during mitosis followed by re-formation at the beginning of a new cell cycle clearly shows that protein concentration cannot be the only determining factor for the formation of Nup214-containing nuclear bodies. It remains to be investigated whether their stability can be regulated by post-translational modifications, e.g. phosphorylation, as they occur during the cell cycle (53).

As observed previously (4), Nup214 fusion proteins inhibit CRM1-dependent nuclear export, most likely as a result of sequestered CRM1 that becomes rate-limiting for the transport reaction. We have now described inhibition of poly(A)⁺ RNA export in cells that overexpress SET-Nup214, as shown previously for full-length Nup214 (37). Similar results were published very recently by Saito et al. (54). The mRNA export factor Tap/Mex67 has been shown to interact with FG regions of Nup214 (55). It remains to be seen whether Tap/Mex67 and/or additional mRNA export factors accumulate in SET-Nup214 nuclear bodies. Our observation that SQSTM1-Nup214 does not lead to a nuclear accumulation of poly(A)⁺ RNA probably relates to the rather low amounts of the fusion protein that end up in the nucleus. Its final concentration is sufficient to inhibit protein but not poly(A)⁺ RNA export.

Potential Role of Nuclear Bodies in Disease

The molecular mechanisms that lead to disease in leukemias containing the Nup214 fusion proteins are unknown. At least two scenarios appear plausible. First, the fusion proteins may inhibit CRM1-dependent protein as well as mRNA export in leukemia cells. We detected co-localization of CRM1 and SET-Nup214 in the leukemia cell line LOUCY (Fig. 2C), suggesting that the corresponding nuclear bodies have overall properties similar to those found in HeLa cells overexpressing the fusion protein. A large number of CRM1 cargo proteins have been identified (30, 56), many of which function as oncogenes or tumor suppressors and could, thus, affect cell proliferation if their physiological nucleocytoplasmic shuttling is affected. Interestingly, selective CRM1 inhibitors are currently being tested with respect to their potential use as anti-cancer drugs (57). Hence, general inhibition of CRM1-dependent export seems to be rather anti-proliferative, suggesting that more specific mechanisms are at play in leukemias involving Nup214 fusions. The inhibition of mRNA export could also affect cell growth, although it is unclear how a general block might have a stimulatory effect. In a second mechanism leading to disease, chromatin-bound CRM1 might recruit Nup214 fusion proteins, as observed previously for Nup98 fusions (51). As a result, a subset of genes might be affected with respect to their expression levels. Interestingly, the leukemogenic CALM-AF10 fusion protein is recruited to chromatin-bound CRM1, suggesting similar mechanisms that might lead to disease (58).

Experimental Procedures

Cell Culture and Transfections

HeLa p4 cells (59) obtained from the NIH AIDS Reagent Program were grown in DMEM (Gibco) supplemented with 10% FCS (Gibco), 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mm l-glutamine (Gibco) at 37 °C and 5% CO₂. LOUCY cells (German Collection of Microorganisms and Cell Cultures (DSMZ) No. ACC-394) were cultured in RPMI 1640 medium (Gibco) supplemented with l-glutamine (Lonza), 20% FBS, and 1% penicillin/streptomycin. MEGAL cells (DSMZ No. ACC-719) were cultured in RPMI 1640 (Gibco) supplemented with 20% FBS and 100 μm 2-mercaptoethanol (Gibco). K562 cells, kindly provided by Alessandro Ori (Jena, Germany), were cultured in RPMI 1640 supplemented with 10% FBS. Plasmids were transfected using the calcium phosphate method (60), Lipofectamine 2000 (Invitrogen), or PolyFect (Qiagen) using up to 0.8 μg of DNA per well of a 24-well plate or 15 μg of DNA/10-cm dish. In some experiments, HeLa cells were treated with 100 μg/ml cycloheximide (Sigma) and/or 10 nm LMB (Enzo Life Sciences) for 1–3 h. LOUCY cells were treated with 20 nm LMB for 3 h.

Immunofluorescence and Confocal Microscopy

HeLa cells were grown on coverslips, fixed with 3.7% formaldehyde in PBS, and permeabilized with 0.5% Triton X-100 in PBS for 5 min. Fixation with 3.5% paraformaldehyde was used for staining with the anti-LAMP2 antibody. For staining with antibodies against Nup88, Nup98, and coilin, cells were fixed with ice-cold 50% methanol, 50% acetone for 3 min at −20 °C without subsequent Triton permeabilization. After incubation with blocking solution (3% BSA in PBS) for 15 min, primary antibodies were added to the blocking solution for 1 h at room temperature. For staining with the anti-RanBP1 antibody, cells were fixed with 3.7% formaldehyde in PBS, and 10% FCS in PBS was used as the blocking solution. Cells were washed three times with blocking solution, and secondary antibodies in blocking solution were added for 1 h at room temperature. After three washes with PBS, the DNA was stained by Hoechst 33258 (1.7 μg/ml) for 3 min followed by three washes with PBS and mounting on microscope slides with FluorSave reagent (Merck Millipore). LOUCY cells were seeded in 24-well plates on coverslips coated with poly-l-lysine (Sigma-Aldrich). After 24–48 h, cells were fixed with 2% formaldehyde in PBS for 15 min at room temperature, permeabilized with 0.2% Triton X-100 in PBS for 10 min, washed with PBS, and blocked with 2% BSA in PBS. Primary antibodies were added, and cells were incubated for 1 h at room temperature. After three washes with PBS, cells were incubated for 1 h with secondary antibodies, washed three times with PBS, and mounted in Mowiol containing DAPI (1 μg/ml). MEGAL and K562 cells were fixed, stained in solution by a standard protocol for nonadherent cells with mild centrifugation after each incubation step, and mounted on microscope slides. Cells were imaged with an LSM 510 Meta confocal microscope (Zeiss) using an LCI Plan-Neofluar ×63 objective (HeLa), a Zeiss LSM 710 confocal laser-scanning microscope (LOUCY), or a Zeiss Axioimager equipped with an Apotome slider and a ×63 NA 1.4 objective (MEGAL and K562 cells). Pictures were not necessarily taken at the same settings but rather to optimize visualization of diffuse versus punctate fluorescence. For quantification of the effect on protein export, the localization of the cargo was assigned to the three categories: N > C (mainly nuclear), N = C (equally distributed between the nucleus and the cytoplasm), and N < C (mainly cytoplasmic). Images were processed and analyzed using ImageJ software and Adobe Photoshop. Figs. 1–8 were assembled with Affinity Designer.

Live Cell Imaging and FRAP

HeLa p4 cells were grown in Nunc Lab-Tek 8-well chambers (Thermo Fisher Scientific) and transfected with pmCherry-SET-Nup214. Live cell imaging was performed at 37 °C with a constant influx of CO₂ in DMEM. A 7-μm Z-stack of 15 planes was collected prior to bleaching. A region of interest (ROI) with a diameter of 60 pixels was defined around an mCherry-SET-Nup214 dot. The ROI was bleached with a high intensity laser pulse using a pixel time of 12.8 μs and 20 iterations. Five additional Z-stacks were acquired after bleaching with a delay of 10 s each. For analysis, the Z-stacks for each time frame were collapsed using ImageJ. Subsequently, in each image, four ROI with a diameter of 3 μm were defined: the bleached dot, a second dot from the same cell, a dot from a non-bleached cell in the same frame, and a background region. The latter two serve as controls. These three dots were tracked in all six time frames, and the position of the ROI was adjusted in each time frame to account for the x/y movement of the dots. The fluorescence intensity of the ROI was measured with ImageJ. The background fluorescence was subtracted from the other ROI, and the intensities were corrected for the general fluorescence loss over time as calculated from the ROI of the non-bleached cell. The fluorescence intensity of the pre-bleached image was set to 100%. Calculations were performed with Microsoft Excel, and the Fig. 8 was assembled with Affinity Designer.

Cell Cycle Arrest

HeLa p4 cells were grown on cover slides in 24-well plates, transfected with mCherry-SET-Nup214 for 24 h, and arrested in G₁/S phase by treatment with 2 mm thymidine in DMEM for 17 h at 37 °C and 5% CO₂. After a 2-h release in DMEM, 100 ng/ml nocodazole was added to arrest the cells in prometaphase. After 12 h at 37 °C and 5% CO₂, cells were washed once with PBS, and DMEM was added to enable release of the cells into mitosis. Slides were removed at various time points during the arrest/release, and cells were fixed with 3.7% formaldehyde. After three washes with PBS, the DNA was stained by Hoechst 33258 (1.7 μg/ml) for 3 min followed by three final washes with PBS and mounting on slides with FluorSave reagent (Merck Millipore).

Quantification of RanBP1 Localization

The fluorescence intensities of the nuclear and cytoplasmic areas of the same size were measured for individual cells using ImageJ. Ratios were calculated in Microsoft Excel, and the mean and standard deviation were plotted with R.

Fluorescence in Situ Hybridization (FISH)

In situ hybridization to detect poly(A)⁺ RNA was described previously (13). Briefly, cells were fixed, permeabilized, and equilibrated in 2× SSC, 25% formamide. Hybridization with 1 ng/ml digoxigenin-labled oligo(dT) was performed for 2 h at 37 °C. RNA was subsequently detected by indirect immunofluorescence with fluorescein-coupled anti-digoxigenin Fab fragments (Roche life Science, catalog No. 11207471910).

Immunoblotting of Cell Lysates

The transfected cells were detached with trypsin and counted. After washing with PBS, cells were dissolved and lysed in sample buffer. The lysates of 30,000 cells were separated by SDS-PAGE followed by immunoblotting and visualization of the transfected proteins with an antibody against RFP.

Subcellular Fractionation

HeLa p4 cells were grown in 24-well plates, transfected with SET-Nup214-myc or SQSTM1-Nup214-myc for 24 h, and treated with 10 nm LMB for 1 h. Cells were detached with trypsin, washed with PBS, and dissolved in PBS at 5000 cells/μl. Lysates were obtained by three cycles of freezing in liquid nitrogen and thawing at 37 °C. Insoluble material was pelleted by centrifugation. The soluble fraction was transferred to a new tube and mixed with SDS loading buffer. The pellet was washed three times with PBS and dissolved in SDS loading buffer. Equal amounts of the pellet (P) or soluble (S) fraction were separated by SDS-PAGE followed by immunoblotting.

Antibodies

Antibodies produced against Nup214 (13) and CRM1 (14) were described previously. LOUCY cells were stained with antibodies against Nup214 (ab40797, Abcam) and CRM1 (catalog No. 611833, BD Biosciences). The anti-coilin antibody was a gift from Angus Lamond (Dundee, UK). Antibodies against Myc (sc-40) and PML (sc-966) as well as mouse IgG (sc-2025) were from Santa Cruz Biotechnology. Anti-LAMP2 (L0668) was from Sigma, anti-RFP (5f8) from ChromoTek, anti-Ran (catalog No. 610340) and anti-RanBP1 (catalog No. 610758) from BD Biosciences, anti-calnexin (ADI-SPA-860) from Enzo Life Sciences, anti-penta-His (catalog No. 34660) from Qiagen, anti-GAPDH (10494-1-AP) from ProteinTech, anti-LC3 (ab48394) from Abcam, and anti-SQSTM1 (catalog No. 7695) from Cell Signaling. Secondary antibodies for immunofluorescence were obtained from Molecular Probes or Life Technologies. Horseradish peroxidase- and fluorophore-coupled secondary antibodies for Western blotting were obtained from Jackson ImmunoResearch and LI-COR Biosciences, respectively.

Plasmids

The pMal-PP expression vector was produced by exchanging the Factor Xa site from pMal-C2 (New England BioLabs) to a recognition site for the PreScission protease by PCR using the oligonucleotides 5′-aaaCCATGGaaaacgcccagaaaggtgaaa-3′ and 5′-tttgaattcGGGCCCCTGGAACAGAACTTCCAGcccgaggttgttgttattgtt-3′.

A SET exon7-Nup214 exon18 fusion as described previously (61) was generated by overlap extension PCR. The N-terminal 810 nucleotides of human SET cDNA were amplified from hSET/pCMV-SPORT6 (clone ID 5587291, Thermo Fisher Scientific) using SET-F-EcoRV (5′-GATATCgccaccatgtcggcgccg-3′) and SET-Nup-R (5′-gatgaaggcgccgaatttcctctccttcctccccttc-3′) primers. The C-terminal 3834 nucleotides of human Nup214 were amplified from a full-length Nup214 plasmid using SET-Nup-F (5′-gaaggggaggaaggagaggaaattcggcgccttcatc-3′) and Nup214-R-XhoI (5′-CTCGAGgcttcgccagccaccaaaac-3′) primers. The amplified SET and Nup214 sequences were used as a template in a second round of amplification using SET-F-EcoRV and Nup214-R-XhoI as primers. The amplified SET-Nup214 fragment was first cloned into the pCR^TM2.1-TOPO vector and then into pcDNA4myc-His (Invitrogen) through EcoRV and XhoI. pcDNA4-SET-myc-His was cloned by amplifying SET from hSET/pCMV-SPORT6 using SET-F-EcoRI (5′-GAATTCgccaccatgtcggcgcc-3′) and SET-R-XhoI (5′-CTCGAGgtcatcttctccttcatcctc-3′) primers and ligating into pcDNA4myc-His (Invitrogen) through EcoRI and XhoI.

For fluorescent protein tagging, the coding sequences for SET-Nup214 and SET were subcloned into pmCherry-C1 by PCR via ApaI/EcoRI using the forward primer 5′-aaaaGAATTCtatgtcggcgccggcgg-3′ and the reverse primers 5′-tttGGGCCCatcagcttcgccagccaccaaaac-3′ and 5′-tttGGGCCCatcagtcatcttctccttcatcctc-3′, respectively. The sequence coding for SQSTM1-Nup214, as described elsewhere (25), was synthesized by GeneArt and subcloned by PCR into pMal-PP via EcoRI/SalI, into pcDNA4-myc-His via Eco32I/ApaI, and into pmCherry-C1 via EcoRI/ApaI using the oligonucleotides 5′-aaaaGAATTCatggcgtcgctcaccgtga-3′/5′-tttGTCGACtcagtgatggtgatggtgatggcttcgccagccaccaaaac-3′, 5′-aaaGATATCatggcgtcgctcaccgtg-3′/5′-aaaGGGCCCgcttcgccagccaccaaaa-3′, and 5′-aaaaGAATTCtatggcgtcgctcaccgtga-3′/5′-tttGGGCCCatcagcttcgccagccaccaaaac-3′, respectively. pEGFP-SQSTM1, a gift from Terje Johansen (Tromsø, Norway), was subcloned into pMal-PP and pmCherry-C1 by PCR via ApaI/EcoRI using the oligonucleotides 5′-tttGAATTCatggcgtcgctcaccgt-3′/5′-tttGTCGACttagtgatggtgatggtgatgcaacggcgggggatgc-3′ and 5′-aaaaGAATTCtatggcgtcgctcaccgtga-3′/5′-tttGGGCCCatcacaacggcgggggatgcttt-3′, respectively. The Nup214 fragments Nup214(1969–2090) and Nup214(813–2090) were cloned into pmRFP-cNLS (14) by PCR via EcoRI/SalI using the forward primers 5′-aaaGAATTCaggcttcggtgctgctc-3′ and 5′-aaaGAATTCagaaattcggcgccttcatc-3′, respectively, and the reverse primer 5′-aaaGTCGACgcttcgccagccaccaa-3′.

EGFP was amplified by PCR using the oligonucleotides 5′-ttttGCTAGCAatggtgagcaagggcgag-3′/5′-ttttGCTAGCgacttgtacagctcgtccat-3′ and inserted into the pEGFP-C1 vector via the NheI site to obtain pdEGFP. The coding sequence of HIV-1 Rev was cloned into pdEGFP by PCR via XhoI and EcoRI using the oligonucleotides 5′-ttttCTCGAGggggaatggcaggaagaagcgga-3′/5′-ttttGAATTCctattctttagctcctgactcca-3′. pEGFP-SPN1 was described previously (62). pFusionRed-ER was from Evrogen.

Constructs with swapped Nup214 portions compared with the original fusion proteins were generated as follows. The plasmid pcDNA4-SET-Nup214(1968–2090)-myc-His was cloned by overlap extension PCR using the primers 5′-gagagtgtggcagctgcccttagccctctgGAAATTCGGCGCCTTCATCA-3′/5′-cttctgagatgagtttttgttcgaagggcccGCTTCGCCAGCCACCAAA-3′ for amplification of the Nup214 part from a plasmid carrying a sequence coding for SQSTM1-Nup214. A sequence coding for SQSTM1 and the N-terminal region of the Nup214 fragment of the SET-Nup214 fusion was synthesized by GeneArt and cloned into the pcDNA4-SET-Nup214-myc-His plasmid using EcoRV/NheI to generate pcDNA4-SQSTM1-Nup214(813-2090)-myc-His.

Co-immunoprecipitation

HeLa p4 cells transfected with plasmids coding for Myc-tagged proteins were grown on 10-cm plates. Cells were scraped off the plates after adding 200 μl of cold lysis buffer/plate (1% Nonidet P-40, 50 mm Tris, pH 8.0, 300 mm NaCl, 5 mm EGTA, 5 mm EDTA, 15 mm MgCl₂, 60 mm β-glycerophosphate, and cOmplete protease inhibitors (Roche life Science)). Cell lysates were transferred to tubes, vortexed, and incubated on ice for 30 min. The total protein content of the clarified lysate (16,100 × g, 4 °C, 20 min) was determined by a BCA protein assay kit (Pierce). 2 mm DTT was added to the lysates, and 0.3–0.5 mg of total protein was used per sample. 1.25 μg of mouse α-Myc antibody or mouse IgG was added to 400-μl reactions containing lysate in the absence or presence of 1 μm RanQ69L-GTP and incubated in a final volume of 400 μl overnight at 4 °C. Protein G-agarose (Roche life Science) was equilibrated overnight at 4 °C in lysis buffer supplemented with 20 mg/ml BSA, added to the samples, and incubated at 4 °C for 4–6 h. The beads were washed three times with lysis buffer, and bound proteins were eluted with SDS sample buffer and analyzed by SDS-PAGE followed by Western blotting.

Expression and Purification of Recombinant Proteins

Expression and purification of Ran (63), CRM1 (64, 65), GST-CRM1 (66), His-Nup214 (14, 35), and MBP-Nup214 (35) were described previously.

MBP-SQSTM1-Nup214-His was expressed in Escherichia coli BL21(DE3) codon+ in MBP-rich medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl, and 0.2% glucose) supplemented with ampicillin. Expression was induced with 100 μm isopropyl β-d-thiogalactopyranoside. Cells were grown overnight at 18 °C, harvested by centrifugation, and lysed in buffer (50 mm Tris, pH 6.8, 300 mm NaCl, 1 mm MgCl₂, 5% glycerol, 4 mm β-mercaptoethanol, 0.1 mm PMSF, and 1 μg/ml each leupeptin, pepstatin, and aprotinin). MBP-SQSTM1-Nup214-His (100,000 × g, 30 min, 4 °C) was purified with nickel-nitrilotriacetic acid-agarose (Qiagen) and amylose resin (New England Biolabs). An NES peptide corresponding to the sequence CVDEMTKKFGTLTIHDTEK was obtained from Pepscan Presto (Lelystad, The Netherlands) and dissolved in 20 mm HEPES-NaOH, pH 7.4, at 1 mm.

Pulldown Assay

50 pmol of GST fusion proteins was immobilized on glutathione-Sepharose (GE Healthcare) equilibrated in pulldown buffer (50 mm Tris, pH 7.4, 200 mm NaCl, 1 mm MgCl₂, 5% glycerol, and 1 mm DTT) supplemented with 20 mg/ml BSA. The immobilized GST fusion proteins were incubated with proteins of interest in 400 μl for 1 h at 4 °C. The beads were washed three times with pulldown buffer. The bound proteins were eluted with SDS sample buffer and analyzed by SDS-PAGE followed by Coomassie staining or Western blotting.

Author Contributions

S. A. P. and R. H. K. designed the study. S. A. P., A. M., C. V., and C. S. prepared the reagents and performed the experiments. S. A. P., A. M., C. V., B. F., C. K., and R. H. K. analyzed the data and prepared the figures. S. A. P., B. F., C. K., and R. H. K. wrote the manuscript.