Pattern of ubiquilin pathology in ALS and FTLD indicates presence of C9ORF72 hexanucleotide expansion

Abstract

C9ORF72-hexanucleotide repeat expansions and ubiquilin-2 (UBQLN2) mutations are recently identified genetic markers in amyotrophic lateral sclerosis (ALS) and frontotemporal lobar degeneration (FTLD). We investigate the relationship between C9ORF72 expansions and the clinical phenotype and neuropathology of ALS and FTLD. Genetic analysis and immunohistochemistry (IHC) were performed on autopsy-confirmed ALS (N = 75), FTLD-TDP (N = 30), AD (N = 14), and controls (N = 11). IHC for neurodegenerative disease pathology consisted of C9ORF72, UBQLN, p62, and TDP-43. A C9ORF72 expansion was identified in 19.4 % of ALS and 31 % of FTLD-TDP cases. ALS cases with C9ORF72 expansions frequently showed a bulbar onset of disease (57 %) and more rapid disease progression to death compared to non-expansion cases. Staining with C9ORF72 antibodies did not yield specific pathology. UBQLN pathology showed a highly distinct pattern in ALS and FTLD-TDP cases with the C9ORF72 expansion, with UBQLN-positive cytoplasmic inclusions in the cerebellar granular layer and extensive UBQLN-positive aggregates and dystrophic neurites in the hippocampal molecular layer and CA regions. These UBQLN pathologies were sufficiently unique to allow correct prediction of cases that were later confirmed to have C9ORF72 expansions by genetic analysis. UBQLN pathology partially co-localized with p62, and to a minor extent with TDP-43 positive dystrophic neurites and spinal cord skein-like inclusions. Our data indicate a pathophysiological link between C9ORF72 expansions and UBQLN proteins in ALS and FTLD-TDP that is associated with a highly characteristic pattern of UBQLN pathology. Our study indicates that this pathology is associated with alterations in clinical phenotype, and suggests that the presence of C9ORF72 repeat expansions may indicate a worse prognosis in ALS.

Fig. 1

Survival in ALS with and without C9ORF72 expansions. Kaplan-Meier plot shows survival of ALS autopsy cases with C9ORF72 repeat expansion (black curve, n = 14) and without C9ORF72 repeat expansion (grey curve, n = 61). Survival time was significantly shorter in cases with C9ORF72 repeat expansion (P = 0.016)

Fig. 2

UBQLN pathology in ALS and FTLD-TDP. a–c Hippocampal dentate gyrus and molecular layer of an ALS case with C9ORF72 expansion, a Dentate gyrus shows cytoplasmic inclusions that are depicted in b in higher resolution, dystrophic neurites with focal swellings as well as irregular, aggregate-like formations, which are depicted in c in higher resolution, d Cerebeliar molecular layer of an ALS case with a C9ORF72 expansion showing cytoplasmic inclusions, which are shown in e in high resolution, f CSC anterior horn α-motoneuron of an ALS case witli a C9ORF72 expansion showing skein-like to more solid cytopiasmic inclusions. Scale bars a–c, e, f 20 μm; d 200 μm

Fig. 3

UBQLN pathology in subgroups of ALS and FTLD-TDP with/without C9ORF72 repeat expansion. The figure illustrates the differentiation of ALS and FTLD-TDP with C9ORF72 expansion (ALS/FTLD-C9+) from non-expansion cases (ALS/FTLD-C9−) by UBQLN pathology. a Hippocampal dentate gyrus and molecular layer of a representative ALS case without C9ORF72 repeat expansion. White dentate gyrus neuronal inclusions are detectable, the molecular layer is free of any pathology (d provides higher resolution image), b, c Extensive UBQLN pathology presenting with dystrophic neurites and aggregate-like formations in the molecular layer of two cases with C9ORF72 repeat expansion (column from b downwards shows an ALS case, column from c downwards shows an FTLD-TDP case). c, f Higher resolution images of aggregate-like formations in the hippocampal molecular layer, g, j Cerebellar granular layer of a FTLD-TDP case without C9ORF72 repeat expansion. The cerebellar granular layer does not show any pathology, h, k Cerebellar granular layer of an ALS case and an FTLD-TDP case (i, l) with C9ORF72 repeat expansion shows cytoplasmic neuronal inclusions which were observed in all ALS/FTLD with C9ORF72 repeat expansion. Scale bars a–c, g–i 200 μm; d, j 50 μm; e, f, k, 1 20 μm

Fig. 4

Extent of UBQLN pathology in ALS and controls. a Bar plot shows mean extent of UBQLN pathology in different regions of the central nervous system in ALS, FTLD-TDP and CTRL, b Bar plot shows mean extent of UBQLN pathology in different regions of the CNS in ALS cases with/without C9ORF72 expansions, c Bar plot shows mean extent of UBQLN pathology in different regions of the CNS in FTLD-TDP case with/without C9ORF72 expansions. Whiskers in bar plot indicate 95 % confidence interval of mean. Amyg amygdala, CSC cervical spinal cord, CA htppocampal CA regions/subiculum, Cer cerebellum, Dent hippocampal dentate gyrus, LSC lumbar spinal cord, MF middle frontal gyrus, Mol hippocampal molecular layer, Mol motor cortex, SMT superior or midtemporal gyrus

Fig. 5

Relation of UBQLN pathology to UBQLN1, TDP-43 and p62. Double-labeling IF analyzed by confocal microscopy shows a hippocampal dystrophic neurites and aggregate-like formations as detected in ALS/FTLD-TDP with C9ORF72 expansion to show a partial co-localization of immunoreactivity for UBQLN and TDP-43 in the aggregate-like alteration, b Higher resolution image shows UBQLN pathology to be surrounded by TDP-43 immunoreactive material. c Hippocampal dentate cytoplasmic neuronal inclusions show immunoreacfivity with a general UBQLN antibody, but not for UBQLN1, d In contrast, hippocampal molecular layer aggregate-like formations show immunoreactivity for general UBQLN and UBQLN1. e Cerebellar granular layer cytoplasmic neuronal inclusions as detected in ALS/FTLD-TDP case with C9ORF72 expansions partially show immunoreactivity for UBQLN and p62. f Hippocampal molecular layer aggregate-like formations only partially show immunoreactivity for both UBQLN and p62, though UBQLN pathology is more extensive than p62 pathology. Scale bars a, c, f 10 μm; b, e 5 μm; d 20 μm