Enrichment of C-terminal fragments in TAR DNA-binding protein-43 cytoplasmic inclusions in brain but not in spinal cord of frontotemporal lobar degeneration and amyotrophic lateral sclerosis

Abstract

TAR DNA-binding protein (TDP-43) has been recently described as a major pathological protein in both frontotemporal dementia with ubiquitin-positive inclusions (FTLD-U) and amyotrophic lateral sclerosis. However, little is known about the relative abundance and distribution of different pathological TDP-43 species, which include hyperphosphorylated, ubiquitinated, and N-terminally cleaved TDP-43. Here, we developed novel N-terminal (N-t) and C-terminal (C-t)-specific TDP-43 antibodies and performed biochemical and immunohistochemical studies to analyze cortical, hippocampal, and spinal cord tissue from frontotemporal dementia with ubiquitin-positive inclusions and amyotrophic lateral sclerosis cases. C-t-specific TDP-43 antibodies revealed similar abundance, morphology, and distribution of dystrophic neurites and neuronal cytoplasmic inclusions in cortex and hippocampus compared with previously described pan-TDP-43 antibodies. By contrast, N-t-specific TDP-43 antibodies only detected a small subset of these lesions. Biochemical studies confirmed the presence of C-t TDP-43 fragments but not extreme N-t fragments. Surprisingly, immunohistochemical analysis of inclusions in spinal cord motor neurons in both diseases showed that they are N-t and C-t positive. TDP-43 inclusions in Alzheimer's disease brains also were examined, and similar enrichment in C-t TDP-43 fragments was observed in cortex and hippocampus. These results show that the composition of the inclusions in brain versus spinal cord tissues differ, with an increased representation of C-t TDP-43 fragments in cortical and hippocampal regions. Therefore, regionally different pathogenic processes may underlie the development of abnormal TDP-43 proteinopathies.

Figure 1

Biochemical characterization of novel termini-specific TDP-43 polyclonal antibodies. A: Schematic diagram of TDP-43 indicating the relative position of the peptides used to generate the pAbs against N-t and C-t regions. The diagram also shows, in scale, the most prominent features of the protein (domains and predicted localization signals). NLS, nuclear localization signal; NES, nuclear export signal; RRM, RNA-recognition motif; Gly-rich, glycine-rich domain. B and C: Representative images showing the immunoblotting pattern for TDP-43 antibodies using T7-tagged recombinant human TDP-43 (C-terminally truncated or full length) (B) or radioimmunoprecipitation assay (RIPA) lysates from QBI293 cells (C). D–F: TDP-43 immunobotting of sarkosyl-insoluble fractions from different FTLD-U subtype cases before (−) and after (+) alkaline phosphatase (AP) treatment. Proteins from frontal cortex of FTLD-U cases were sequentially extracted with buffers of increasing strength and subjected to SDS-polyacrylamide gel electrophoresis. Dephosphorylation of FTLD-U urea extracts with AP followed by immunoblotting with pan TDP-43 (D) and C-t TDP-43 pAb (E) collapsed the 45-kDa band into the 43-kDa band and separated TDP-43 fragments into four immunoreactive ∼23- to ∼27-kDa bands. Immunoblotting with N-t TDP-43 pAb (F) shows the collapse of the 45-kDa band, but no evidence for low molecular weight species. This indicates that pathological, sarkosyl-insoluble TDP-43 is abnormally phosphorylated and that low molecular weight pathological bands contain extreme C-t TDP-43 fragments.

Figure 2

Anti-TDP-43 C-t antibodies detect more neuropathology than N-t antibodies in dentate gyrus and frontal cortex of FTLD-U subtypes. A-F: Adjacent sections of hippocampal tissue from FTLD-U cases show robust labeling of neuronal cytoplasmic inclusions (arrowheads) in the dentate gyrus (DG) using both pan TDP-43 antibody (A and D) and C-t TDP-43 pAb (B and E), in contrast to marginal staining with N-t TDP-43 pAb (C and F). D–F: High-power views of the fields depicted in A–C. Note clearing of nuclear TDP-43 (arrows) in inclusion-bearing neurons compared with that of nonaffected neurons (*). G–O: Adjacent sections of frontal cortex (FC) in FTLD-U cases show the characteristic features of type I pathology (G–I), with long and tortuous dystrophic neurites with relatively few neuronal cytoplasmic inclusions and no neuronal intranuclear inclusions. Type II cases (J–L) present numerous cytoplasmic inclusions and infrequent neuritic profiles. Type III cases (M–O) have numerous neuronal cytoplasmic inclusions and dystrophic neurites and occasional intranuclear inclusions in lamina II. Note the similar immunoreactivity pattern of characteristic TDP-43 pathology, ie, cytoplasmic inclusions (arrowheads) and dystrophic neurites, for pan TDP-43 pAb (G, J, and M) and C-t TDP-43 pAb (H, K, and N). Use of N-t TDP-43 pAb (I, L, and O) revealed remarkably reduced staining for both types of pathology (arrows) but robust nuclear staining of normal cells. Insets in J–O show a high-power detail of representative staining. Scale bars: 25 μm (A–C and G–O); 7 μm (D–F and insets in J–O).

Figure 3

Colocalization studies of C-t and N-t TDP-43 antibodies in FTLD-U. Double-label immunofluorescence of FTLD-U cases with a C-t TDP-43 mAb (A, D, G, and J) and N-t TDP-43 pAb (B, E, H, and K) shows almost complete lack of N-t pAb staining for the C-t-positive inclusions in the presence of normal nuclear TDP-43 staining (Merge + DAPI panels: C, F, I, and L). This was observed in frontal cortices from all FTLD-U types analyzed (A–I) and, to a lesser extent, in hippocampal granule cells (J–L). Scale bars: 50 μm (A–L); 5 μm (insets).

Figure 4

Single-color immunohistochemistry and double-label immunofluorescence of neuronal cytoplasmic inclusions in ALS spinal cord. A–C: Adjacent sections of spinal cord in sporadic ALS cases show that the same motor neurons display inclusions positive for all three antibodies tested, pan TDP-43 (A), C-t TDP-43 (B), and N-t TDP-43 pAb (C). Note positive nuclear staining for all three antibodies (insets). A high-power view of one motor neuron (arrow) from A–C is shown (A′–C′). D–I: Double-label immunofluorescence shows that both filamentous and compact, round cytoplasmic inclusions in sporadic ALS cases display complete colocalization (F and I) of C-t TDP-43 mAb (D and G) and N-t TDP-43 pAb (E and H) staining. For comparison, the same cases were stained with a combination of commercial TDP-43 mAb (J and M) and C-t TDP-43 pAb (K and N), which also show complete colocalization (L and O). Scale bars: 50 μm (A–C); 20 μm (A′–C′ and D–O).

Figure 5

TDP-43 pathology in FTLD-U with MND spinal cord and ALS frontal cortex. A set of three adjacent sections of ALS frontal cortex (A–C) or FTLD-U with MND spinal cord tissue (D–F) were stained with pan TDP-43 (A and D), C-t TDP-43 (B and E), or N-t TDP-43 pAb (C and F). Whereas the pathology staining pattern for FTLD-U with MND spinal cord was similar for all three antibodies (D–F), the cortical pathology in ALS cases with dementia showed reduced N-t TDP-43 immunoreactivity compared with the other two antibodies (arrowheads; A–C). High-power images are shown in D′–F′ and insets on A–C. Scale bars: 20 μm (A–C and D′–F′); 50 μm (D–F); 6 μm (insets in A–C).

Figure 6

Semiquantitative analysis of TDP-43-positive pathology in FTLD-U and ALS cases reveal differential enrichment of C-t fragments in brain regions. Scatter plots illustrating individual values and means of the neuropathological immunoreactivity scores for pan-TDP-43 (crosses), C-t TDP-43 (dots), and N-t TDP-43 (triangles) polyclonal antibodies. A: TDP-43 pathology score values from FTLD-U cases in frontal cortex gray matter (left), hippocampal dentate gyrus (middle), and lower motor neurons (LMN) in spinal cord (right). B: Comparable data sets compiled from FTLD-U with MND cases. C: Data from ALS cases. *P < 0.05; **P < 0.01; ***P < 0.001 (Dunn’s multiple comparison post test).

Figure 7

Differential C-t and N-t anti-TDP-43 immunoreactivities in AD brain. A–F: Adjacent sections of hippocampus (A–C) or frontal cortex (FC; D–F) from AD brains show robust labeling of neuronal cytoplasmic inclusions in the dentate gyrus (DG) and dystrophic neurites in FC using both pan TDP-43 antibody (A and D) and C-t TDP-43 pAb (B and E); in contrast, only marginal pathological staining is revealed with N-t TDP-43 pAb (C and F). High-power images corresponding to the boxed regions are shown in insets in A–F. Scale bars: 25 μm (A–F); 12 μm (all insets). G: Scatter plots showing individual values and means of the neuropathological immunoreactivity scores for pan-TDP-43 (crosses), C-t TDP-43 (dots), and N-t TDP-43 (triangles) pAbs. The diagrams display TDP-43 pathology score values from AD cases in hippocampal dentate gyrus (left) and frontal cortex gray matter (right). *P < 0.05; **P < 0.01 (Dunn’s multiple comparison post test). H: Summary of demographic characteristics of AD patients used in this study. F, female: M, male; No, number.