A cellular model of TDP-43 induces phosphorylated TDP-43 aggregation with distinct changes in solubility and autophagy dysregulation

Abstract

Amyotrophic lateral sclerosis (ALS) is an incurable neurodegenerative disease that affects neurons in the brain and spinal cord, causing loss of muscle control, and eventually leads to death. Phosphorylated transactive response DNA binding protein-43 (TDP-43) is the major pathological protein in both sporadic and familial ALS, forming cytoplasmic aggregates in over 95% of cases. Of the 10-15% of ALS cases that are familial, mutations in TDP-43 represent about 5% of those with a family history. We have developed an in vitro overexpression model by introducing three familial ALS mutations (A315T, M337V, and S379P) in the TDP-43 (TARDBP) gene which we define as 3X-TDP-43. This overexpression model TDP-43 shows deficits in autophagy flux and colocalization of TDP-43 with stress granules. We also observe a progressive shift of TDP-43 to the cytoplasm in this model. This overexpression model shows a reduction in solubility of phosphorylated TDP-43 from RIPA to urea soluble. Four glycolytic enzymes, phosphoglycerate kinase one (PGK1), aldolase A (ALDOA), enolase 1 (ENO1), and pyruvate dehydrogenase kinase 1 (PDK1) show significant time-dependent decreases in 3X-TDP-43 expressing cells. Shotgun proteomic analysis shows global changes in the importin subunit alpha-1 (KPNA2), heat shock 70 kDa protein 1A (HSPA1A), and protein disulfide-isomerase A3 (PDIA3) expression levels and coimmunoprecipitation reveals that these proteins complex with TDP-43. Overall, these results suggest that the 3X-TDP-43 model may provide new insights into pathophysiology and an avenue for drug screening in vitro for those suffering from ALS and related TDP-43 proteinopathies.

Keywords: ALS; TDP‐43; aggregation; autophagy; glycolysis; stress granules.

Fig. 1

Overexpression of TDP‐43 with three familial mutations results in cytoplasmic aggregation. (A) Domain structure of TDP‐43: nuclear localization signal (NLS), RNA recognition motif 1 (RRM1), RNA recognition motif 2 (RRM2), nuclear export signal (NES) (new evidence suggests export through passive diffusion), and glycine‐rich region. Three familial ALS mutations (A315T, M337V, and S379P) were introduced into human TDP‐43 cDNA cloned into pcDNA 3.1 (46–48). (B) Immunofluorescence staining probed for TDP‐43 C‐terminal (555 nm, cyan) and DAPI (358 nm, blue) from HEK293T transfected from pcDNA‐GFP (magenta) and 3X‐TDP‐43‐GFP (observed in over 3 independent experiments).

Fig. 2

Overexpression of 3X‐TDP‐43 has greater TDP C‐terminal fragments compared to single mutant TDP‐43. (A) Western blot probed for TDP‐43 C‐terminal and β‐Actin from HEK293T cells transfected with TDP‐43 A315T, TDP‐43 M337V, TDP‐43 S379P, and 3X‐TDP‐43 for 72 h. (B–D) Quantification of TDP‐43 C‐terminal/β‐actin from western blots at 43, 35, and 25 kDa. One‐way Anova with error bars representing standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 (n = 3 biological replicates, experiment performed 3 separate times).

Fig. 3

Neuroblastoma (SH‐SY5Y) cells transfected with 3XTDP‐43‐GFP at 72 h show cytoplasmic aggregation. Immunofluorescence staining probed for TDP‐43 C‐terminal (555 nm, cyan) and DAPI (358 nm, blue) from SH‐SY5Y cells transfected with pcDNA‐GFP (magenta) and 3X‐TDP‐43‐GFP (magenta) for 72 h (observed in over 3 independent experiments).

Fig. 4

Overexpression of 3X‐TDP‐43 at 24‐h incubation results in cytoplasmic accumulation and C‐terminal fragments. (A) Western blot of TDP‐43 C‐terminal, Histone H3, and tubulin from nuclear(N)/cytoplasmic(C) fractionation of HEK293T transfected with 3XTDP‐43 and CMV‐tTA (control) for 24 h. (B–F) Quantification of nuclear and cytoplasmic fraction from western blot at 24 h. (Quantification‐ ratio of arbitrary units, cytoplasmic‐TDP‐43/tubulin; nuclear‐TDP‐43/H3). Unpaired t‐test with error bars representing standard deviation. **P < 0.01, ***P < 0.001. (G) Immunofluorescence staining probed for TDP‐43 C‐terminal (488 nm, green) and DAPI (358 nm, blue) from HEK293T cells transfected with CMV‐tTA and 3X‐TDP‐43 for 24 h (n = 3 biological replicates, experiment performed 6 separate times).

Fig. 5

Overexpression of 3X‐TDP‐43 at 48‐h incubation results in cytoplasmic accumulation and C‐terminal fragments. (A) Western blot of TDP‐43 C‐terminal, Histone H3, and tubulin from nuclear(N)/cytoplasmic(C) fractionation of HEK293T transfected with 3XTDP‐43 and CMV‐tTA (control) for 48 h. (B–F) Quantification of nuclear and cytoplasmic fraction from western blot at 48 h. (Quantification‐ratio of arbitrary units, cytoplasmic‐TDP‐43/tubulin; nuclear‐TDP‐43/H3). Unpaired t‐test with error bars representing standard deviation. *P < 0.05, **P < 0.01, ****P < 0.0001. (G) Immunofluorescence staining probed for TDP‐43 C‐terminal (488 nm, green) and DAPI (358 nm, blue) from HEK293T transfected from CMV‐tTA and 3X‐TDP‐43 for 48‐h incubation (n = 3 biological replicates, experiment performed 6 separate times).

Fig. 6

Overexpression of 3X‐TDP‐43 at 72‐h incubation results in cytoplasmic accumulation and C‐terminal fragments. (A) Western blot of TDP‐43 C‐terminal, Histone H3, and tubulin from nuclear(N)/cytoplasmic(C) fractionation of HEK293T transfected with 3XTDP‐43 and CMV‐tTA (control) for 72 h. (B–H) Quantification of nuclear and cytoplasmic fraction from western blot at 72 h. (Quantification‐ratio of arbitrary units, cytoplasmic‐TDP‐43/tubulin; nuclear‐TDP‐43/H3). Unpaired t‐test with error bars representing standard deviation. **P < 0.01, ***P < 0.001, ****P < 0.0001. (I) Immunofluorescence staining probed for TDP‐43 C‐terminal (488 nm, green) and DAPI (358 nm, blue) from HEK293T transfected from CMV‐tTA and 3X‐TDP‐43 for 72‐h incubation (n = 3 biological replicates, experiment performed 6 separate times with similar results).

Fig. 7

In cells expressing 3X‐TDP‐43 for 72 h the nuclear volume was significantly increased. (A) SH‐SY5Y cells transfected with pcDNA‐GFP and 3X‐TDP‐43‐GFP for 72 h probed with DAPI (358 nm, blue). White outlines measure the area of the nuclei. (B) Quantification of nuclear area. Unpaired t‐test, two tailed with error bars represent standard deviation. *P < 0.05 (n = 3 biological replicates, experiment performed 3 separate times with similar results).

Fig. 8

Overexpression of 3X‐TDP‐43 results in changes in TDP‐43 C‐terminal fragments at different timepoints. (A) Western blot of whole cell lysate from CMV‐tTA (72 h), 3X‐TDP43 (24, 48, 72 h) transfected HEK293T cells. (B–E) Quantification of western blot. One‐way Anova with error bars representing standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (n = 3 biological replicates, experiment performed 3 separate times with similar results). Tubulin control from same protein lysate as Fig. 11.

Fig. 9

Overexpression of 3X‐TDP‐43 displays time‐dependent differential cleavage products. (A) Western blot probed for TDP‐43 C‐terminal and tubulin from lysed HEK293T cells transfected with CMV‐tTA for 48 h and 3X‐TDP‐43 for 48 and 72 h. (B–M) Quantification of TDP‐43 C‐terminal/tubulin from western blot at ~125, ~55, ~43, ~42, ~40, ~38, ~37, ~35, ~29, ~25, ~24, and ~20 kDa. One‐way ANOVA with Tukey's correction for Type I error and error bars represent standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (n = 3 biological replicates, experiment performed 3 separate times with similar results).

Fig. 10

Overexpression of 3X‐TDP‐43 at 72 h compared to human TDP‐43‐V5 increases expression of 125 kDa upper fragments (A) Western blots probed for TDP‐43 C‐terminal and tubulin at long and short exposure from lysed HEK293T cells transfected with TDP‐43 V5 and 3X‐TDP‐43 for 72 h. (B–D) Quantification of TDP‐43 C‐terminal/tubulin from western blots at 125, 35, and 25 kDa. Unpaired t‐test, two tailed with error bars represent standard deviation. **P < 0.01, ****P < 0.0001 (n = 3 biological replicates, experiment performed 3 separate times with similar results).

Fig. 11

Overexpression of 3X‐TDP‐43 induces autophagy dysfunction and fluxes at different timepoints. (A) Western blot probed for SQSTM1 (p62), LC3B and tubulin from HEK293T lysates transfected with CMV‐tTA (72 h) and 3X‐TDP‐43 for 24, 48, and 72 h. (B, C) Quantification of p62/tubulin and LC3B/tubulin from western blot. One‐way Anova with error bars representing standard deviation. *P < 0.05, **P < 0.01 (n = 3 biological replicates, experiment performed 3 separate times with similar results). Tubulin control from same protein lysate as Fig. 8.

Fig. 12

3X‐TDP‐43 colocalizes with TIAR at 72 h. (A) Immunofluorescence staining probed and TIAR (555 nm, cyan, ProteinTech Group) from SHSY5Y cells transfected with pcDNA‐GFP (magenta, control) and 3X‐TDP‐43‐GFP. (B) Western blot probed for TIAR and tubulin from lysed SH‐SY5Y cells transfected with CMV‐tTA and 3X‐TDP‐43. (C) Quantification of TIAR/tubulin from western blot. Unpaired t‐test with error bars representing standard deviation. No significant difference, P > 0.05 (n = 3 biological replicates, experiment performed 3 separate times with similar results).

Fig. 13

Overexpression of 3X‐TDP‐43 displays time‐dependent reduction in glycolysis enzyme protein levels. (A) Western blot probed for ADOA, PGK1, PDHK1, ENO1 normalized to β‐Actin from lysed HEK293T cells transfected with CMV‐tTA and 3X‐TDP‐43 for 24, 48, and 72 h. (B–E) Quantification of Aldolase A/β‐Actin, PGK1/β‐Actin, PDHK1/β‐Actin, and Enolase 1/β‐Actin. One‐way Anova with error bars representing standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001. Non‐significant‐ ns, P > 0.05 (n = 3 biological replicates, experiment performed 6 separate times with similar results).

Fig. 14

The glycolic enzyme PGK1 protein decreased in HEK293 cells expressing 3X‐TDP‐43. (A) Western blot probed for PGK1 and β‐Actin from lysed HEK293T cells transfected with CMV‐tTA and 3X‐TDP‐43 for 24 h. (B) Quantification of PGK1/β‐Actin. Unpaired T‐test with error bars representing standard deviation. ****P < 0.0001 (n = 6 biological replicates, experiment performed 3 separate times with similar results).

Fig. 15

Mutant TDP‐43 overexpression induces aggregation of phosphorylated TDP‐43 in HEK293T cells. (A) Western blot of urea‐soluble fraction (U) and RIPA soluble samples (R) probed for pTDP‐43 (S409/S410) from lysed HEK293T cells transfected with CMV‐tTA and 3XTDP‐43 for 72 h. (B) Immunofluorescence staining probed for pTDP‐43 (S409/S410) (488 nm, magenta, PTG) and TDP‐43 C‐terminal (555 nm, cyan) and DAPI (358 nm, blue) from HEK293T cells transfected with CMV‐tTA (control) and 3X‐TDP‐43 for 72 h. 1000× magnification. (C) Quantification of cytoplasmic pTDP‐43 (S409/410) from immunofluorescence. Unpaired T‐test with error bars representing standard deviation. ****P < 0.0001. (D) Immunofluorescence staining for pTDP‐43 (S409/S410) (555 nm, cyan, PTG) and DAPI (358 nm, blue) from HEK293T cells transfected with CMV‐tTA (control) and 3X‐TDP‐43 for 72 h. 1000× magnification (n = 3 biological replicates, experiment performed 3 separate times with similar results).

Fig. 16

Phosphorylated TDP‐43 aggregation in neuroblastoma (SH‐SY5Y) cells expressing 3X‐TDP‐43‐GFP. Immunofluorescence staining probed for pTDP43 (S409/S410) (555 nm, cyan, Proteintech), DAPI (358 nm, blue), and 3X‐TDP‐43‐GFP (magenta) compared with pcDNA‐GFP (magenta) control. 1000X magnification (experiment performed 3 separate times with similar results).

Fig. 17

PDIA3 and KPNA2 interact in complex with TDP‐43 and KPNA2 and are significantly increased in the nucleus of HEK293T overexpressing 3X‐TDP‐43. (A) Western blot of immunoprecipitation using human parahippocampal gyrus (ADNC, F, age 91, brain weight 1208) pulled down with IgG mouse (IgGM), IgG rabbit (IgGR), KNA2, PDIA3, and TDP‐43 C‐terminal and probed for PDIA3, KPNA2, and TDP‐43 C‐terminal. (B) Western blot probed for KPNA2, Histone H2B, Histone H3, and tubulin from nuclear/cytoplasmic fractionation of HEK293T transfected with 3X‐TDP‐43 and CMV‐tTA (control) for 72 h. (C, D) Quantification of KPNA2/histone H3, and KPNA2/tubulin. (E) Nuclear/Cytoplasmic Ratio for KPNA2 quantification. (F) Quantification of Histone 2B/Histone H3. Unpaired T‐test with error bars representing standard deviation. *P < 0.05 (proteomics experiment performed once with n = 3 biological replicates, Western n = 3 biological replicates with 3 separate experiments with similar results and IP validated in human brain through 3 separate experiments with similar results).

Fig. 18

3X‐TDP‐43 overexpression induces reduction in PDIA3 and increase in KPNA2. (A) Western blot of HEK293T transfected with CMV‐tTA for 72 h and 3X‐TDP‐43 at 24‐, 48‐, and 72‐h timepoints and probed for HSPA1A, PDIA3, KPNA2, and tubulin. (B–D) Quantification of HSPA1A, PDIA3, and KPNA2. One‐way Anova with error bars representing standard deviation. *P < 0.05, **P < 0.01, ***P < 0.001 (n = 3 biological replicates with 3 separate experiments with similar results).