Dipeptide repeat protein inclusions are rare in the spinal cord and almost absent from motor neurons in C9ORF72 mutant amyotrophic lateral sclerosis and are unlikely to cause their degeneration

Abstract

Introduction: Cytoplasmic TDP-43 inclusions are the pathological hallmark of amyotrophic lateral sclerosis (ALS) and tau-negative frontotemporal lobar dementia (FTLD). The G4C2 repeat mutation in C9ORF72 is the most common cause of ALS and FTLD in which, in addition to TDP-43 inclusions, five different di-peptide repeat (DPR) proteins have been identified. Di-peptide repeat proteins are translated in a non-canonical fashion from sense and antisense transcripts of the G4C2 repeat (GP, GA, GR, PA, PR). DPR inclusions are abundant in the cerebellum, as well as in the frontal and temporal lobes of ALS and FTLD patients and some are neurotoxic in a range of cellular and animal models, implying that DPR aggregation directly contributes to disease pathogenesis. Here we sought to quantify inclusions for each DPR and TDP-43 in ALS cases with and without the C9ORF72 mutation. We characterised the abundance of DPRs and their cellular location and compared this to cytoplasmic TDP-43 inclusions in order to explore the role of each inclusion in lower motor neuron degeneration.

Results: Spinal cord sections from ten cases positive for the C9ORF72 repeat expansion (ALS-C9+ve) and five cases that were not were probed by double immunofluorescence staining for individual DPRs and TDP-43. Inclusions immunoreactive for each of the DPRs were present in the spinal cord but they were rare or very rare in abundance (in descending order of frequency: GA, GP, GR, PA and PR). TDP-43 cytoplasmic inclusions were 45- to 750-fold more frequent than any DPR, and fewer than 4 % of DPR inclusions colocalized with TDP-43 inclusions. In motor neurons, a single cytoplasmic DPR inclusion was detected (0.1 %) in contrast to the 34 % of motor neurons that contained cytoplasmic TDP-43 inclusions. Furthermore, the number of TDP-43 inclusions in ALS cases with and without the C9ORF72 mutation was nearly identical.

Conclusions: For all other neurodegenerative diseases, the neurotoxic protein aggregates are detected in the affected population of neurons. TDP-43 cytoplasmic aggregation is the dominant feature of ALS spinal cord pathology irrespective of C9ORF72 mutation status. The near absence of DPR inclusions in spinal cord motor neurons challenges their contribution to lower motor neuron degeneration in ALS-C9+ve cases.

Fig. 1

Representative images of DPR inclusions. a-j are representative immunofluorescence confocal images of DPR (red) and TDP-43 (green) (scale bar = 20 μm). a-e show cytoplasmic DPR inclusions, which are often perinuclear. f-h show nuclear DPR inclusions. i and j show evidence of DPR inclusions colocalizing with TDP-43. k and l are lower power images of the anterior horn showing the relative abundance of DPR and TDP-43 inclusions (scale bar = 50 μm). l shows a poly-GA inclusion, which is not in the motor neurons (dotted nucleus) that show extensive cytoplasmic TDP deposition. k shows a poly-GP inclusion within a motor neuron (identified by its large pale nucleus), which does not show cytoplasmic TDP-43 inclusions. Motor neurons were identified by their characteristic large pale nucleus (dotted line)

Fig. 2

Frequency of TDP-43 and DPR inclusion per case, per section. The number of TDP-43 and DPR immunoreactive inclusions was counted for each case and normalised per spinal cord section. The graph shows the number of TDP-43 and the five different aggregates in order of abundance. There are many more TDP-43 aggregates than any DPR. Note: The Y axis scale is logarithmic scale. Horizontal bars show mean ± SEM