OUP user menu

Nuclear import impairment causes cytoplasmic trans-activation response DNA-binding protein accumulation and is associated with frontotemporal lobar degeneration

Agnes L. Nishimura, Vera Župunski, Claire Troakes, Claudia Kathe, Pietro Fratta, Michael Howell, Jean–Marc Gallo, Tibor Hortobágyi, Christopher E. Shaw, Boris Rogelj
DOI: http://dx.doi.org/10.1093/brain/awq111 1763-1771 First published online: 14 May 2010


Trans-activation response DNA-binding protein (TDP-43) accumulation is the major component of ubiquitinated protein inclusions found in patients with amyotrophic lateral sclerosis, and frontotemporal lobar degeneration with TDP-43 positive ubiquitinated inclusions, recently relabelled the ‘TDP-43 proteinopathies’. TDP-43 is predominantly located in the nucleus, however, in disease it mislocalizes to the cytoplasm where it aggregates to form hallmark pathological inclusions. The identification of TDP-43 mutations in familial and sporadic amyotrophic lateral sclerosis cases confirms its pathogenic role; but it is wild-type TDP-43 that is deposited in the vast majority of TDP-43 proteinopathies, implicating other unknown factors for its mislocalization and aggregation. One such mechanism may be defective nuclear import of TDP-43 protein, as a disruption of its nuclear localization signal leads to mislocalization and aggregation of TDP-43 in the cytoplasm. In order to explore the factors that regulate the nuclear import of TDP-43, we used a small interfering RNA library to silence 82 proteins involved in nuclear transport and found that knockdowns of karyopherin-β1 and cellular apoptosis susceptibility protein resulted in marked cytoplasmic accumulation of TDP-43. In glutathione S-transferase pull-down assays, TDP-43 bound to karyopherin-αs, thereby confirming the classical nuclear import pathway for the import of TDP-43. Analysis of the expression of chosen nuclear import factors in post-mortem brain samples from patients with TDP-43 positive frontotemporal lobar degeneration, and spinal cord samples from patients with amyotrophic lateral sclerosis, revealed a considerable reduction in expression of cellular apoptosis susceptibility protein in frontotemporal lobar degeneration. We propose that cellular apoptosis susceptibility protein associated defective nuclear transport may play a mechanistic role in the pathogenesis of the TDP-43 positive frontotemporal lobar degeneration.

  • TDP-43
  • frontotemporal lobar degeneration
  • nuclear transport
  • karyopherin-β1
  • cellular apoptosis susceptibility


Frontotemporal lobar degeneration with trans-activation response DNA-binding protein (TDP-43) positive ubiquitinated inclusions (FTLD–TDP) and amyotrophic lateral sclerosis (ALS) are incurable and fatal neurodegenerative disorders with overlapping clinical, genetic and pathological features, which are now referred to as TDP-43 proteinopathies (Neumann et al., 2006; Kwong et al., 2008; Mackenzie et al., 2009). The common pathological hallmark is the aggregation of TDP-43 protein to form ubiquitinated inclusions. These are most striking in the cytoplasm of affected neurons, often accompanied by loss of TDP-43 staining from the nucleus. It has been proposed that loss of normal nuclear function and/or toxic gain of function due to cytoplasmic accumulation and aggregation may be the underlying mechanism of cellular dysfunction and death in this group of disorders (Lagier-Tourenne and Cleveland, 2009).

Our group, and subsequently others, have identified mutations in the gene encoding TDP-43 (TARDBP) in 1–4% of familial and sporadic ALS cases (Sreedharan et al., 2008; Lagier-Tourenne and Cleveland, 2009) and recently frontotemporal dementia (Benajiba et al., 2009). However, the vast majority of patients do not have genomic TARDBP mutations and the reasons why TDP-43 accumulates and aggregates in the cytoplasm are unknown. One cause may be defective nuclear import; the nuclear transport machinery itself may be defective and/or post-translational modification of TDP-43 may make the protein unable to bind effectively to the nuclear transport machinery.

TDP-43 is a ubiquitously expressed nuclear protein that binds to DNA and RNA and has a function in regulating the transcription and RNA processing (Wang et al., 2004; Ayala et al., 2006; Buratti and Baralle, 2008). It has two RNA recognition motifs, a nuclear export site, a nuclear localization signal and a glycine-rich C-terminal domain.

The transfer of TDP-43 from cytoplasm into the nucleus is an energy-dependent process that utilizes a bipartite nuclear localization signal in the N-terminal portion of TDP-43 (Ayala et al., 2008; Winton et al., 2008a). Replacement of basic residues in nuclear localization signal promotes a cytoplasmic localization of TDP-43 and formation of large nuclear and cytoplasmic aggregates.

In classical nuclear import a member of the karyopherin-α family recognizes and binds to the nuclear localization signal of the cytoplasmic cargo protein. Karyopherin-β1 then binds to karyopherin-α and the karyopherin-β1–karyopherin-α-cargo complex is actively transported through the nuclear pore complex into the nucleus (Lange et al., 2007; Sorokin et al., 2007; Terry et al., 2007). Most cargo proteins have a preference for a specific karyopherin-α; others can be imported by several or all karyopherin-αs with similar efficacy (Kohler et al., 1999; Ma and Cao, 2006). Within the nucleus, dissociation of the karyopherin-β1/karyopherin-α/cargo complex is triggered by binding of the small GTPase Ran to karyopherin-β1. Unbound karyopherin-β1 is then recycled directly back to the cytoplasm, while karyopherin-αs are returned to the cytoplasm via another member of the karyopherin-β family—cellular apoptosis susceptibility (CAS) (Kutay et al., 1997).

Here we demonstrate that down-regulation of members of the classical nuclear import pathway leads to accumulation of TDP-43 in the cytoplasm. We also demonstrate that CAS is significantly reduced in the brains of patients with FTLD–TDP and propose a mechanistic link between defective nuclear transport and cytoplasmic accumulation in FTLD–TDP.

Materials and methods


Primary antibodies used in this study are as follows: rabbit polyclonal anti-TDP-43 (Proteintech), mouse monoclonal anti-TDP-43 (Abnova), mouse monoclonal anti-fused in sarcoma (Santa Cruz), mouse monoclonal anti-glyceraldehyde 3-phosphate dehydrogenase (Sigma), rabbit polyclonal anti–histone H3 (Cell Signalling Technology), mouse monoclonal anti-CAS (Santa Cruz), mouse anti-karyopherin-β1 (Sigma), mouse monoclonal anti-karyopherin-α2 (Sigma), mouse monoclonal anti-karyopherin-α1 (Abnova), goat polyclonal anti-karyopherin-α6 (Abnova), rabbit polyclonal anti-ubiquitin (Dako), mouse monoclonal anti-p62 (BD Transduction Laboratories) and rabbit polyclonal anti-green fluorescent protein (Cell Signalling Technology). The secondary antibodies used were goat anti-mouse and anti-rabbit Alexa Fluor 488 and 568 (Invitrogen) for immunocytochemistry and biotinylated swine anti-rabbit (DAKO) for immunohistochemistry.

Plasmid constructs

pDONR221 plasmids with karyopherin-α1 (BC002374), karyopherin-α2 (BC005978), karyopherin-α3 (BC017355), karyopherin-α4 (BC034493), karyopherin-α6 (BC020520), karyopherin-β1 (BC003572) and transportin 1 (BC040340) were obtained from PlasmID (Dana-Farber/Harvard Cancer Centre DNA Resource Core). The clones were then transferred to pDEST-GST (pDEST with N-terminal glutathione S-transferase derived from pGEX-2TK) using the Gateway system (Invitrogen). pEGFP vector containing wild-type TDP-43 has been described previously (Sreedharan et al., 2008). TDP-43 with deleted nuclear localization signal (TDP-43ΔNLS) was created by deleting the bipartite nuclear localization signal (amino acids 82–98) using QuickChange II kit (Stratagene) and primers (GTCAACTATCCAAAAGATAACGCAGTCCAGAAAACATCCG and CGGATGTTTTCTGGACTGCGTTATCTTTTGGATAGTTGAC).

Small interfering RNA

Custom made SMARTPool human small interfering RNA library (Dharmacon) against 82 proteins involved in nuclear transport was used for initial screening (Supplementary Table 1). Confirmation studies for selected targets were done with human ON-TARGETplus (Dharmacon) and mouse and human Stealth RNAi (Invitrogen, San Diego, CA) (Supplementary Tables 2 and 3).

Cell culture and transfection

Human neuroblastoma SHSY-5Y, mouse neuroblastoma N2a and human embryonic kidney 293T cells were cultured in Dulbecco’s modified Eagle’s medium-glutamax, 10% foetal calf serum (Invitrogen) at 5% CO2 and 37°C. Primary cortical neurons were obtained from E-16 mouse embryos (Williamson et al., 2008). Neurons were cultured on plates coated with poly-d-lysine (Sigma) in medium containing Neurobasal, B27 and l-glutamine (Invitrogen) at 5% CO2 and 37°C. Initial screening was done in Optilux Microtest 96-well assay plates (BD Biosciences). Confirmation studies were carried out on 13 mm glass coverslips in 24-well plates. Transfections were performed using Lipofectamine 2000 and Lipofectamine RNAiMAX (Invitrogen), following the manufacturer’s protocol for small interfering RNA. Cells were harvested 48–96 h later.

Immunocytochemical analyses

For immunofluorescence, cells were washed with phosphate buffered saline and fixed with 4% paraformaldehyde for 15 min, followed by permeabilization with 0.5% Triton X-100 in phosphate buffered saline for another 15 min. Subsequently the coverslips were washed with phosphate buffered saline and blocked with blocking buffer (10% foetal bovine serum in phosphate buffered saline) for 1 h. Cells were then incubated with primary antibody diluted in blocking buffer at room temperature for one hour or overnight at 4°C, followed by incubation with secondary antibody diluted in blocking buffer for another hour. They were then washed in phosphate buffered saline and the nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Cells transfected with karyopherin-β1, CAS, nucleoporin-like 1 (NUPL1), nucleoporin-62 kD (NUP62) and nucleoporin-54 kDa (NUP54) were stained with anti-TDP-43 and counted (cytoplasmic TDP-43 versus total) (n > 1000 cells were counted from three independent experiments). Statistical analysis was performed using t-test; P-values below 0.05 were considered significant.

Cell fractionation and western blot

Cells were washed with cold phosphate buffered saline and harvested with 10 mM Tris pH7.4, 10 mM NaCl, 3 mM MgCl2, 1 mM EDTA, 1 mM EGTA and complete protease inhibitor cocktail (Roche Applied Science). After 10 min incubation on ice, the lysate was centrifuged at 375g for 5 min at 4°C. The supernatant (cytoplasmic fraction) was centrifuged for another 30 min at 150 000g and 4°C and the pellet (nuclear fraction) was washed three times with the solution mentioned above, plus 0.1% NP-40, and centrifuged for 5 min at 800g and 4°C. Proteins were quantified using non-interfering Protein Assay Kit (Calbiochem) and loaded in NuPAGE Novex 10% Bis–Tris pre-cast gels (Invitrogen). The membranes were blocked with 10% non-fat dry milk for at least 1 h and incubated with primary antibodies overnight at 4°C. After serial washes with 1% Tween-20 and Tris-buffered saline, the membranes were incubated with secondary antibody; followed by washes in Tween-20 and Tris-buffered saline, incubated in chemiluminescence reagent (ECL; Immobilon Western, Millipore) and exposed to film (Amersham hyperfilm ECL, GE Healthcare Limited, UK). Data were presented as mean ± SEM values. Student’s t-test was used for comparing two group means. A P-value of <0.05 was considered statistically significant.

Pull-down assays

Karyopherin-αs, karyopherin-β1 and transportin 1 glutathione S-transferase-fusion proteins were prepared according to the protocol described previously (Kesavapany et al., 2001). Interaction with endogenous TDP-43 was analysed in SHSY-5Y cells, while specificity of the nuclear localization signal was analysed in human embryonic kidney 293T cells transiently transfected with pEGFP-TDP-43 and pEGFP-TDP-NLS. Cells were lysed with radioimmunoprecipitation assay buffer [1% Triton X-100, 0.5% sodium deoxycholoate, 20 mM Tris pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, protease inhibitor cocktail (Roche) and 0.1% RNase A (Ambion) and T1 (Roche)] and cleared by centrifugation at 13 000 rpm at 4°C, for 30 min. RNases were added to remove RNA-mediated interactions that are found in heterogeneous nuclear ribonucleoprotein complexes (Choi and Dreyfuss, 1984; Fukuda et al., 2009). Protein lysates were incubated for 3 h with different glutathione S-transferase fusion sepharose slurries at 4°C in rotation; washed three times in radioimmunoprecipitation assay buffer and boiled with sodium dodecyl sulphate polyacrylamide gel electrophoresis sample buffer. Western blots were performed as described above.

Patients’ data and characterization of post-mortem brain samples

Ten percent formalin-fixed, paraffin-embedded tissue blocks and fresh frozen samples of temporal lobe sections containing the superior and middle temporal gyri (Brodmann areas 20/21) and cervical or thoracic spinal cord were available at the MRC London Neurodegenerative Diseases Brain Bank (Institute of Psychiatry, King’s College London, UK). Brain samples were obtained from 11 patients with FTLD–TDP (three females and eight males; mean ± SD age 64.9 ± 11.49 years; mean ± SD post-mortem interval 19.3 ± 11.6 h) and nine matched (age, sex, post-mortem interval and pH) normal controls (five females and four males; mean ± SD age 63 ± 9.3 years; mean ± SD post-mortem interval 27.9 ± 4.8 h) (Supplementary Table 4). Spinal cord samples were obtained from 10 patients with ALS (six females and four males; mean ± SD age 64.6 ± 10.2 years; mean ± SD post-mortem interval 32.6 ± 20.3 h) and 10 matched (age, sex, post-mortem interval and pH) normal controls (three females and seven males; mean ± SD age 70.5 ± 17.3 years; mean ± SD post-mortem interval 45.5 ± 24.4 h) (Supplementary Table 6). Controls were defined as subject with no clinical history and no neuropathological evidence of a neurodegenerative condition. The consent for autopsy, neuropathological assessment and research use were obtained from all subjects in accordance with local and national Research Ethics Committee approved donation. Block taking for histological and immunohistochemical studies and neuropathological assessment of FTLD–TDP and control cases was performed in accordance with published guidelines (Cairns et al., 2007).

Fresh-frozen post-mortem tissue was homogenized in 10 v/w of high salt buffer [100 mM 2-(N-morpholino) ethane sulphonic acid (MES) (pH 7.4), 0.5 mM MgCl2, 1 mM EGTA, 1 M NaCl, 50 mM imidazole, protease inhibitor cocktail (Roche)]. The homogenate was mixed with 2× sodium dodecyl sulphate polyacrylamide gel electrophoresis loading buffer and boiled for 10 min. Samples were centrifuged for 20 min at 13 000 rpm and 4°C. Equal volumes of samples were loaded on 26-well NuPAGE Novex 10% Bis–Tris pre-cast gels (Invitrogen). Western blots were preformed as described above.

For immunohistochemistry, 7 μm thick sections representing the superior and medial temporal gyrus or cervical or thoracic spinal cord were cut from paraffin-embedded tissue blocks. Immunohistochemical staining was carried out according to a standard protocol for paraffin-embedded tissue sections (Maekawa et al., 2004). Briefly, sections were microwaved in citrate buffer to enhance antigen retrieval. All primary antibodies (CAS, karyopherin-α1, karyopherin-α2 and karyopherin-α6) were used at a concentration of 1:150 (overnight incubation at 4°C). Sections were developed using 3,3′-diaminobenzidine and counter-stained with haematoxylin. To ensure standardized conditions, sections were stained as a single batch and incubations times were identical.

Immunostaining (location and intensity) was examined by a consultant neuropathologist and the intensity of respective nuclear and cytoplasmic staining scored (blind to diagnosis) using the following standard semiquantitative scale: 0 = no staining; 1 = mild staining; 2 = moderate staining; 3 = intense staining.


Disruption of the classical nuclear import pathway leads to cytoplasmic accumulation of TDP-43

In order to identify the factors involved in the transport of TDP-43 into the nucleus, we screened a custom made SMARTpool human small interfering RNA library that targets 82 nuclear transport proteins (Supplementary Table 1).

Five nuclear transport factors consistently showed significant accumulation of TDP-43 in the cytoplasm of human neuroblastoma SHSY-5Y cells: NUPL1, NUP62, NUP54, CAS and karyopherin-β1 (Fig. 1A; Supplementary Fig. 1). None of these factors had any effect on distribution of fused in sarcoma (FUS), which was used as a control nuclear protein with a different nuclear import pathway (Lee et al., 2006). NUPL1, NUP62 and NUP54 are an intrinsic and essential part of the nuclear pore complex and their knockdown would be expected to have a generalized effect on nuclear protein import, including that of TDP-43. We therefore explored the role of CAS and karyopherin-β1, as they might play a more selective role in the nuclear import of TDP-43.

Figure 1

Screening of nuclear transport small interfering RNA library for proteins involved in nuclear import of TDP-43. (A) Percentage of cells containing positive cytoplasmic staining when karyopherin-β1 (si KPNB1), CAS (si CAS), NUPL1 (si NUPL1), NUP62 (si NUP62) and NUP54 (si NUP54) are knockdown in SHSY-5Y cells. Immunostaining of TDP-43 (green) and DAPI (blue) in (B) SHSY-5Y and (C) mouse cortical neurons. RNA interference of karyopherin-β1 and CAS show cytoplasmic accumulation of TDP-43. Higher magnification shows that TDP-43 forms aggregates in the cytoplasm. si = small interfering.

The involvement of CAS and karyopherin-β1 in localization of TDP-43 in SHSY-5Y cells was confirmed using two additional sets of small interfering RNAs (Fig. 1B, Supplementary Tables 2 and 3, Supplementary Fig. 2). With one of the sets we also repeated RNA interference for Ran-binding protein 5, Importin-7 and Transportin 1 and again saw no change in TDP-43 localization (data not shown). Furthermore, we observed formation of cytoplasmic TDP-43 aggregates following the knockdowns of karyopherin-β1 and CAS (Fig. 1B). To characterize these aggregates further, we co-stained these cells with ubiquitin and p62 and neither showed co-localization with TDP-43 aggregates (Supplementary Fig. 3). RNA interference of karyopherin-β1 and CAS in mouse neuroblastoma N2a cells (Supplementary Fig. 4 and Supplementary Table 3) revealed similar results to those observed in SHSY-5Y cells and western blot analysis of cytoplasmic and nuclear fractions confirmed that RNA interference of karyopherin-β1 and CAS increases TDP-43 in the cytoplasmic fraction (Fig. 2A). The cytoplasmic mislocalization of TDP-43 following karyopherin-β1 and CAS downregulation was also observed in primary mouse cortical neurons, further confirming that loss of CAS or karyopherin-β1 leads to cytoplasmic accumulation of TDP-43 (Fig. 1C).

Figure 2

RNA interference of karyopherin-β1 and CAS affect TDP-43 and karyopherin-α2 distribution. (A) TDP-43 accumulates in the cytoplasm when karyopherin-β1 and CAS are silenced and karyopherin-α accumulates in the nuclear fraction when CAS is silenced. Western blot analysis of TDP-43 and karyopherin-α2 in cytoplasmic and nuclear fractions of N2a cells following knockdowns of karyopherin-β1 and CAS. Membranes were probed to detect TDP-43, karyopherin-α2 (KPNA2), Histone H3 (H3) as loading control for nuclear fraction and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as loading control for cytoplasmic fraction (bottom). (B) Change in karyopherin-α distribution correlates with cytoplasmic TDP-43 accumulation. Immunostaining of TDP-43 (green), karyopherin-α2 (red) and DAPI (blue) in SHSY-5Y cells. RNA interference of karyopherin-β1 (si KPNB1) and CAS induce TDP-43 accumulation in the cytoplasm. RNA interference of karyopherin-β1 induces mislocalization of karyopherin-α2 in the nucleus whereas RNA interference of CAS induces karyopherin-α2 accumulation in the cytoplasm. (C) Western blot of glutathione S-transferase–karyopherin-α pull-downs probing for endogenous TDP-43. Glutathione S-transferase–karyopherin-α1, glutathione S-transferase (GST)–karyopherin-α2, glutathione S-transferase–karyopherin-α3, glutathione S-transferase–karyopherin-α4, glutathione S-transferase–karyopherin-α6, glutathione S-transferase–karyopherin-β1 and glutathione S–transferase–transportin1 fusion proteins were incubated with SHSY-5Y cell lysate. Sepharose only and glutathione S–transferase were used as non-specific binding controls. Western blot of the pull-down was probed with TDP-43 monoclonal antibody. (D) Karyopherin-α binds to the transfected with enhanced green fluorescent protein tagged TDP-43 or TDP-43 with deleted nuclear localization signal and western blot was probed with anti-green fluorescent protein antibody. Glutathione S–transferase only was used as non-specific binding control.

The effect of karyopherin-β1 and CAS knockdown strongly implicates a role for karyopherin-α proteins in the import of TDP-43, therefore we first analysed their effect on distribution of karyopherin-αs (Fig. 2A, B). Karyopherin-αs are normally distributed equally in between the nucleus and cytoplasm. Following karyopherin-β1 knockdown, there was a reduction in the level of nuclear karyopherin-αs. Conversely, when CAS is silenced, karyopherin-α levels in the nucleus are increased. We therefore repeated the RNA interference of all six karyopherin-αs with another set of small interfering RNAs and again found no effect on TDP-43 distribution (data not shown). To explore if more than one karyopherin-α is involved in binding to TDP-43, we used a glutathione S-transferase-pull-down method to analyse the direct interaction of TDP-43 with karyopherin-α proteins. Incubation of SHSY-5Y cell lysate with glutathione S-transferase–karyopherin-α fusion proteins revealed that endogenous TDP-43 bound to all of the karyopherin-αs (Fig. 2C). We also observed TDP-43 binding to glutathione S-transferase–karyopherin-β1, which may be mediated by karyopherin-αs. Endogenous TDP-43 did not bind to glutathione S-transferase-transportin 1 suggesting that the binding of TDP-43 to karyopherin-α is specific, as transportin 1 is known to bind to a different type of nuclear localization signal (Lee et al., 2006). To define further the binding of TDP-43 to karyopherin-α via the nuclear localization signal of TDP-43, we repeated the glutathione S-transferase-pull-down experiments with glutathione S-transferase–karyopherin-α fusion proteins using cell lysates containing overexpressed wild-type TDP-43 or TDP-43ΔNLS. As expected, overexpressed wild-type TDP-43 localized in the nucleus and TDP-43ΔNLS showed cytoplasmic localization (Supplementary Fig. 5). Overexpressed wild-type TDP-43 bound to glutathione S-transferase–karyopherin-α and TDP-43ΔNLS did not (Fig. 2D).

Reduction of nuclear import factors in post-mortem brains

To investigate whether the levels of nuclear transport factors are altered in TDP-43 proteinopathies, we analysed their abundance and distribution in the affected cortical regions of patients with TDP-43 positive FTLD (FTLD–TDP) (Supplementary Table 4) and spinal cords from patients with ALS (Supplementary Table 6). Western blots of total protein lysates from medial temporal lobe samples were probed with antibodies to CAS, karyopherin-β1, karyopherin-α1, karyopherin-α2 and karyopherin-α6. CAS was below the detection threshold in 10/11 patients with FTLD–TDP but was readily detected in all controls studied (P = 1.03 × 10–5, t-test) (Fig. 3A, B). A significant reduction was also observed for karyopherin-α2 (P = 0.02) and no significant change for karyopherin-β1, karyopherin-α1 and karyopherin-α6.

Figure 3

CAS and karyopherin-α2 are reduced in brain samples from patients with FTLD–TDP compared to controls. (A) Western blot was performed on cortical lysates from patients with FTLD–TDP and control brain samples. Membranes were probed to detect CAS, karyopherin-β1 (KPNB1), karyopherin-α1 (KPNA1), karyopherin-α2 (KPNA2), karyopherin-α6 (KPNA6) and glyceraldehyde 3-phosphate dehydrogenase as a loading control (bottom). (B) Quantification showing significant loss of CAS and karyopherin-α2 in patients with FTLD–TDP in comparison to controls [***P < 0.001 and *P < 0.05, t-test, data represent ± SEM from n = 9 (controls) and n = 11 (FTLD–TDP)]. (C) Immunohistochemistry showing decreased nuclear and cytoplasmic CAS and karyopherin-α2 labelling in the temporal lobe cortex in FTLD–TDP. Arrows correspond to cytoplasmic staining of CAS and karyopherin-α2 in controls. Scale bars represent 50 µm.

We then analysed the expression of CAS and karyopherin-α2 by scoring their immunoreactivity in temporal lobe sections (Fig. 3C). Both proteins were observed predominantly in neurons in cortical layers 3 and 5. Consistent with the western blot data, CAS immunoreactivity was reduced in the nucleus and cytoplasm in FTLD–TDP cases compared to controls (Fig. 3C and Supplementary Table 5). For karyopherin-α2, nuclear staining was reduced in FTLD–TDP compared to controls. Cytoplasmic staining of karyopherin-α2 was detected at low levels in controls but absent in FTLD–TDP. Thus, both proteins involved in the nuclear import of TDP-43 are reduced in affected neurons in FTLD–TDP cases.

Western blot analysis of CAS expression in spinal cord lysates from ALS cases and controls was below the detection threshold using the protocol for detecting CAS in the brain. Only after a 10-fold increase in the concentration of the primary antibody (1:500) and long exposure times was a signal detected, however no significant difference was observed (Fig. 4A, Supplementary Fig. 6). A significant increase was observed for karyopherin-α2 (P = 8.8 × 10–5) and decrease for karyopherin-α6 (P = 0.006) and no significant change for karyopherin-β1 and karyopherin-α1 (Fig. 4A and B). CAS could not be detected using the immunohistochemical analysis. In view of western blotting results, we concluded that this protein was below the detection threshold for immunohistochemistry. Karyopherin-α2 immunostaining showed an increase in cytoplasmic staining in ALS samples compared to controls (Fig. 4C and Supplementary Table 7).

Figure 4

Karyopherin-α2 and karyopherin-α6 change their expression in spinal cord samples from TDP-positive sporadic ALS patients compared to controls. (A) Western blot was performed on spinal cord lysates from patients and control brain samples. Membranes were probed to detect CAS (top), karyopherin-β1, karyopherin-α1, karyopherin-α2, karyopherin-α6 and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) as a loading control (bottom). (B) Quantification showing significant increase of karyopherin-α2 and decrease karyopherin-α6 in patients with ALS in comparison to controls [**P < 0.01 and ***P < 0.001, t-test, data represent ± SEM from n = 10 (controls) and n = 10 (ALS)]. (C) Immunohistochemistry showing increased cytoplasmic karyopherin-α2 labelling in the cervical or thoracic spinal cord in sporadic ALS. Arrows correspond to cytoplasmic staining of karyopherin-α2 in controls. Scale bars represent 50 µm.


Cytoplasmic mislocalization and aggregation of TDP-43 is the hallmark pathology of ALS and frontotemporal lobar degeneration with Ubiquitin-positive inclusions and there is increasing evidence that this event is mechanistic in initiating neurodegeneration. Neurons with cytoplasmic TDP-43 inclusions have very little nuclear staining, implying that defective nuclear import may contribute to its mislocalization and aggregation. Here we have demonstrated that disruption of the classical nuclear import pathway in neurons leads to the cytoplasmic accumulation of TDP-43 (Fig. 2D). Furthermore, we show that brains of patients with sporadic FTLD–TDP have a large reduction in expression of an important nuclear transport factor, i.e. CAS (Fig. 3). This was not observed in the spinal cord samples of patients with ALS, suggesting that CAS-associated defective nuclear transport may play a mechanistic role in the pathogenesis of FTLD–TDP.

It has been shown that the nuclear import of TDP-43 is an energy-dependent process that utilizes a bipartite nuclear localization signal in the N-terminal portion (amino acids 82–98) of TDP-43 (Winton et al., 2008a) and that A90V mutation in the nuclear localization signal disrupts the localization of TDP-43 (Winton et al., 2008b). The results of our RNA interference screen, followed by two confirmation knockdowns of karyopherin-β1 and CAS using additional different sets of small interfering RNAs, indicated that the members of the karyopherin-β1 pathway are involved in the process of TDP-43 import and that their disruption leads to accumulation of TDP-43 in the cytoplasm (Fig. 5). Overall this effect was observed in three different neuronal cell types of human and mouse origin (SHSY-5Y, N2a and primary mouse neurons). The knockdown of karyopherin-β1 caused karyopherin-αs to remain in the cytoplasm and knockdown of CAS caused karyopherin-αs to accumulate in the nucleus, which agrees with the proposed mechanism of their transport across the nuclear pore complex (Kutay et al., 1997; Lange et al., 2007; Sorokin et al., 2007; Terry et al., 2007). This loss of karyopherin-αs from the cytoplasm, or the inhibition of their import into the nucleus, may lead to cytoplasmic accumulation of TDP-43. However, two attempts at RNA interference of individual karyopherin-αs that did not show any retention of TDP-43 in the cytoplasm suggested that TDP-43 may be imported into the nucleus by more than one karyopherin-α, as was observed for some other cargo proteins (Kohler et al., 1999; Ma and Cao, 2006). Glutathione S-transferase–karyopherin-α pull-down showed that TDP-43 binds to all karyopherin-αs studied, giving additional confirmation that TDP-43 is transported into the nucleus via the classical nuclear import pathway and that it binds to karyopherin-αs in a redundant manner. The specificity of the interaction of TDP-43 with karyopherin-αs was further confirmed by lack of binding to the TDP-43ΔNLS.

Figure 5

Proposed mechanism of TDP-43 import into the nucleus. Knockdown of any of the five illustrated nuclear import proteins will lead to reduction of nuclear import of TDP-43. KPNA = karyopherin-α; KPNB1 = karyopherin-β1.

Following the CAS or karyopherin-β1 knockdowns, we also observed formation of TDP-43 immunopositive aggregates. Similar observations were made when TDP-43 nuclear localization signal was mutated (Ayala et al., 2008; Winton et al., 2008a). We did not observe any colocalization with p62 and ubiquitin (Supplementary Fig. 3), suggesting that the aggregates were not akin to the ones observed in patient motor neurons or that additional factors were necessary for their formation in cells. Indeed, overexpression of the TDP-43 with mutated nuclear localization signal in combination with proteasome inhibitor MG132 has been necessary to cause ubiquitinated p62-positive aggregates in cells (Nonaka et al., 2009). The effect of overexpression of wild-type TDP-43 on aggregate formation has been recapitulated in a recent transgenic mouse model (Wils et al., 2010). However, it has also recently been observed that aggregates may not be necessary for the cellular toxicity associated with cytoplasmic overaccumulation of TDP-43 (Barmada et al., 2010).

We have also demonstrated that levels of CAS and, to a lesser extent, karyopherin-α2 are reduced in the brains of patients with FTLD–TDP. Such decrease in concentration of proteins transporting TDP-43 into the nucleus might contribute to its pathological mislocalization and aggregation.

The lack of difference in CAS expression in ALS could be due to overall lower expression of CAS in the spinal cord tissue, nevertheless it suggests that another mechanism may be responsible for cytoplasmic TDP-43 accumulation in ALS. Nuclear contour irregularity and loss of nuclear karyopherin-β1 distribution in anterior horn cells in ALS has been observed very recently (Kinoshita et al., 2009), suggesting that abnormalities in nuclear transport may also be associated with ALS. Karyopherin-α2 was decreased in FTLD–TDP samples and increased in ALS. The karyopherin-α2 increase in ALS could be due to the cytoplasmic increase, which we observed in immunohistochemical sections (Supplementary Table 7) and the nuclear contour irregularity observed by Kinoshita et al. (2009) may be one of the causes for such redistribution.

Nuclear transport is impaired with age (Pujol et al., 2002; D'Angelo et al., 2009), which is a major independent risk factor for neurodegenerative disorders including FTLD (Yankner et al., 2008). Age-dependent decreases in the protein concentrations of karyopherin-α2 and CAS have been reported in fibroblasts unaccompanied by change in the level of karyopherin-β1 (Pujol et al., 2002), which correlates with our observations for FTLD–TDP. Karyopherin-αs are also mislocalized in cells exposed to oxidative stress (Kodiha et al., 2008), another factor implicated it the pathogenesis of FTLD (Martinez et al., 2009).

We therefore propose a possible mechanism for TDP-43 proteinopathies where impaired efficiency of TDP-43 import into the nucleus leads to its accumulation in the cytoplasm. This mislocalzation of TDP-43 may result in a loss of its nuclear function or cause a toxic gain of function due to its overaccumulation and aggregation in the cytoplasm. Collectively, our results provide evidence that defective nuclear transport may play a mechanistic role in the pathogenesis of TDP-43 proteinopathies.


Medical Research Council (MRC); Motor Neurone Disease Association (MNDA); Psychiatry Research Trust; Cancer Research UK (CRUK); NIHR Biomedical Research Centre for Mental Health at the South London and Maudsley NHS Foundation Trust; Slovenian Research Agency.

Supplementary material

Supplementary material is available at Brain online.


The authors thank Ms Sophie Morris for administrative support; Dr Wendy Noble and Ms Alessia Usardi for providing primary mouse cortical neurons and Ms. Sofia Luz for assistance in the laboratory work.


  • *These authors contributed equally to this work.

  • Abbreviations:
    amyotrophic lateral sclerosis
    cellular apoptosis susceptibility
    frontotemporal lobar degeneration with TDP-43 positive ubiquitinated inclusions
    nucleoporin-like 1
    nucleoporin-62 kD
    nucleoporin-54 kD
    trans-activation response DNA-binding protein


View Abstract