Nemo-like kinase disrupts nuclear import and drives TDP43 mislocalization in ALS

Abstract

Cytoplasmic transactive response DNA-binding protein 43 (TDP43) mislocalization and aggregation are pathological hallmarks of amyotrophic lateral sclerosis (ALS). However, the initial cellular insults that lead to TDP43 mislocalization remain unclear. In this study, we demonstrate that nemo-like kinase (NLK) - a proline-directed serine-threonine kinase - promotes the mislocalization of TDP43 and other RNA-binding proteins by disrupting nuclear import. NLK levels were selectively elevated in neurons exhibiting TDP43 mislocalization in tissues from patients with ALS, and genetic reduction of NLK reduced toxicity in human neuron models of ALS. Our findings suggest that NLK is a promising therapeutic target for neurodegenerative diseases.

Keywords: ALS; Cell biology; Molecular pathology; Neuroscience; Transport.

Figure 1. Overexpression of NLK leads to cytoplasmic accumulation of TDP43.

(A) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT, followed by immunofluorescence using antibodies against FLAG (green) and TDP43 (magenta); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (B–E) Superplots showing the nuclear/cytoplasmic (N/C) ratio (B), whole-cell intensity (C), nuclear intensity (D), and cytoplasmic intensity (E) of TDP43 from cells shown in A. (F) Scatterplot showing TDP43 N/C ratio as a function of FLAG-NLK-WT whole-cell intensity. Data are shown as the mean ± SEM. ***P < 0.0001, unpaired 2-tailed t test with Welch’s correction.

Figure 2. Overexpression of NLK leads to cytoplasmic accumulation of ALS/FTLD–relevant RNA-binding proteins.

(A–F) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT, followed by immunofluorescence using antibodies against FLAG (green), TDP43 (magenta), and FUS (A), HNRNPA2B1 (C), or MATR3 (E) (cyan); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (B, D, and F) Superplots of the N/C ratio of FUS (A), HNRNPA2B1 (C), or MATR3 (E) in cells overexpressing KN or WT NLK. Data are shown as the mean ± SEM. **P < 0.01, ***P < 0.0001, unpaired 2-tailed t test with Welch’s correction.

Figure 3. NLK overexpression disrupts NLS-dependent nuclear import.

(A–H) HEK293T cells were cotransfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT and either of the following: NLS reporter eYFP-NLS^TDP43 (A and B), eYFP-NLS^FUS (C and D), eYFP-NLS^MATR3 (E and F), or eYFP-NLS^SV40 (G and H). Representative images of immunofluorescence are shown in A, C, E, and G, using antibodies against FLAG (green) and either TDP43, FUS, or Matrin-3 (magenta); direct fluorescence of reporter fusion protein is shown in yellow. DNA was stained with Hoechst (blue). Scale bar: 10 μm. Superplots of quantification of NLS-reporter localization nuclear/cytoplasmic ratio (N/C) are shown in B, D, F, and H. Data are shown as the mean ± SEM. **P < 0.01, ****P < 0.0001, unpaired 2-tailed t test with Welch’s correction.

Figure 4. NLK-dependent mislocalization of TDP43 does not depend on nuclear accumulation of KPNA2 and KPNB1.

(A–D) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT. Representative images of immunofluorescence are shown in A and C, using antibodies against FLAG (green), TDP43 (magenta), and KPNB1 (A) or KPNA2 (C) (cyan); DNA was stained with Hoechst (blue). Scale bar: 5 μm. Superplots of quantification of N/C ratio of the indicated proteins are shown in B and D. Data are shown as the mean ± SEM. Statistical significance indicated as such in all panels: ***P < 0.001 (unpaired 2-tailed t test with Welch’s correction). (E–H) HEK293T cells were cotransfected with plasmids encoding FLAG-NLK-WT and either mApple (negative control) or V5-FBXW7. Representative images of immunofluorescence are shown in E and G using antibodies against FLAG (green), V5 (magenta), and KPNA2 (E) or TDP43 (G) (cyan).Direct fluorescence of mApple is shown in magenta. DNA was stained with Hoechst (blue). In E and G, arrows indicate cells coexpressing FLAG-NLK WT and V5-FBXW7, while asterisks indicate cells expressing FLAG-NLK WT only. Scale bar: 20 μm. Superplots of N/C ratios of the indicated proteins are shown in F and H. Data are shown as the mean ± SEM. Statistical significance indicated as such in all panels: ***P < 0.001 (unpaired 2-tailed t test with Welch’s correction).

Figure 5. NLK interacts with the RanBP2-RanGAP1 complex.

(A) Representative Western blots from HEK293T cells transfected with either empty vector (–) or FLAG-NLK-WT (+). Molecular weights in kDa are indicated on the left. (B) Western blot analysis after IP of lysates from empty vector (–) or FLAG-NLK-WT–expressing cells using the indicated antibodies. For KPNB1, the arrow indicates the KPNB1-reactive band, and the asterisk (*) indicates SUMO–RanGAP1. (C) Higher-resolution Western blot after IP of RanBP2 from empty vector (–) or FLAG-NLK-WT–expressing cells.

Figure 6. NLK overexpression mislocalizes RanBP2-RanGAP1 and disrupts the Ran gradient.

(A–H) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT, followed by immunofluorescence using antibodies against FLAG (green), TDP43 (magenta), and either RanGAP1 (A), RanBP2 (C), MAb414 (FG-nucleoporins; E), or Ran (G) (cyan); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (B, D, F, and H) Superplots of nuclear rim-to-cytoplasmic ratio (Nuc. Rim/Cyto) of RanGAP1 (A), RanBP2 (C), and MAb414 (E) or nuclear to cytoplasmic ratio (N/C) of Ran (G). Data are shown as the mean ± SEM. Statistical significance is indicated as follows in all panels: **P < 0.01, ***P < 0.001. P values calculated using unpaired 2-tailed t test with Welch’s correction.

Figure 7. NLK-induced TDP43 mislocalization is RNA independent, but NLK overexpression affects RNA distribution.

(A) HEK293T cells were cotransfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT and a GFP-tagged RNA-binding mutant TDP43 (TDP43 F147/9L; F2L), followed by immunofluorescence using antibodies against FLAG (green) and FUS (magenta) and direct visualization of tagged protein (cyan); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (B) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT, followed by immunofluorescence using antibodies against FLAG (green), TDP43 (magenta), and NXF1 (cyan); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (C) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT, followed by polyA FISH (cyan) and immunofluorescence for FLAG (green) and TDP43 (magenta); DNA was stained with Hoechst (blue). Scale bar: 10 μm. (D–F) Quantification of data presented in A–C. (D) Percentage of cells with cytoplasmic GFP TDP F2L signal in cells expressing FLAG-NLK-KN or FLAG-NLK-WT. (E and F) Superplots of N/C ratio of NXF1 or PolyA FISH in HEK293T cells expressing FLAG-NLK-KN or FLAG-NLK-WT. Data are shown as mean ± SD. *P < 0.05, ****P < 0.0001, unpaired 2-tailed t test with Welch’s correction.

Figure 8. NLK overexpression drives dissolution of nuclear speckles.

(A–C) HEK293T cells were transfected with plasmids encoding either FLAG-NLK-KN or FLAG-NLK-WT. (A–C) Representative images of immunofluorescence for FLAG (green), TDP43 (magenta), and markers of speckles (SC-35; cyan) (A), paraspeckles (SFPQ; cyan) (B), or nucleolus (nucleophosmin, Npm; cyan) (C). DNA is stained with Hoechst (blue). Scale bar: 10 μm. (D–F) Superplots of number of speckles (D), paraspeckles (E), or nucleoli (F) in HEK293T cells expressing FLAG-NLK-KN or FLAG-NLK-WT. Data are shown as mean ± SD. **P < 0.01, unpaired 2-tailed t test with Welch’s correction.

Figure 9. NLK overexpression disrupts nuclear import in primary rat neurons.

(A–E) Primary rodent cortical neurons were transfected with either SNAP-FLAG (SF; negative control) or SNAP-FLAG-NLK (SF-NLK), followed by immunofluorescence using antibodies against FLAG (green) and TDP43, FUS (magenta), Ran, RanBP2, RanGAP1, or MAb414 (cyan); DNA was stained with Hoechst (blue). Scale bar: 10 μm.

Figure 10. NLK overexpression is toxic in primary rat neurons.

(A and B) Primary rodent cortical neurons were cotransfected with plasmids encoding either SF or SF-NLK and EGFP (survival marker), treated with SNAP-647 dye at T1 to visualize SNAP-positive neurons, and tracked by longitudinal microscopy to determine neuronal fate. Scale bar: 20 μm. (C) Cumulative hazard plot showing the relative risk of death in neurons expressing either SF or SF-NLK. HR = 1.616. ***P = 1.494 × 10^–59. (D) Cox proportional hazards model predicting relative hazard based on SNAP-FLAG-NLK expression intensity. Solid lines represent estimated hazard; color gradients reflect expression levels: gray (low), light red (medium), and red (high). Dashed lines represent 95% CI.

Figure 11. TDP43 pathology is associated with NLK overexpression in human model systems and patient samples.

(A) Schematic of generation of mbOrgs. (B) Normalized NLK counts from RNA-Seq performed on WT or GRN^–/– mbOrgs. Data are shown as mean ± SD. **P < 0.01, unpaired 2-tailed t test with Welch’s correction. (C) qRT-PCR analysis of NLK mRNA levels in WT and GRN^–/– mbOrgs (2 biological replicates, 3 technical replicates per condition). Superplot of NLK expression normalized to GAPDH. Line = mean; error bars = standard deviation. *P < 0.05, unpaired 2-tailed t test with Welch’s correction. (D) NLK normalized counts from RNA-Seq performed on TDP43-positive and -negative nuclei. Data are shown as mean ± SD. ***P < 0.001, paired 2-tailed t test with Welch’s correction. (E) Dual IHC for NLK and TDP43, performed on spinal cord tissue from 4 patients with sporadic ALS. Images deconvolved using Fiji. Scale bar for upper panels: 100 μm. Scale bar for lower panels: 20 μm.

Figure 12. Genetic NLK reduction prevents neurodegeneration in human neuron ALS/FTLD models.

(A) qRT-PCR for NLK mRNA in iPSC-derived neurons transduced with lentivirus encoding either nontargeting shRNA (shNT) or NLK-targeting shRNA (shNLK). Data are shown as the mean ± SEM. ***P < 0.001, unpaired 2-tailed t test with Welch’s correction. (B and C) Isogenic WT and TDP43 M337V iPSC-derived neurons were transduced with lentivirus encoding either shNT or shNLK and tracked by longitudinal microscopy to assess neuronal survival. Scale bar: 20 μm. (D) Cumulative hazard plot showing the relative risk of death in ND and C9 neurons expressing either shNT or shNLK. HR = 0.40, ***P < 0.001. (E) Cumulative hazard plot showing the relative risk of death in WT and TDP43 M337V neurons expressing either shNT or shNLK. HR = 0.23, ***P = 0.0013.