Abstract Body

ALS and FTD are fatal and incurable neurodegenerative diseases, characterized by accumulation of pathologic forms of RNA-binding proteins, predominantly TDP-43 and FUS. The functional consequences and potential neurotoxic effects of these FUS or TDP-43 aggregates observed in postmortem brains are debated. Combining structural protein analysis with cellular systems we uncovered a novel and unexpected mechanism that counteracts pathologic aggregation of TDP-43 (1). We also recently discovered that FTD heterogeneity is associated with alternate pathological TDP-43 conformations, reminiscent of prion strains (2). Moreover, we recently showed that cytoplasmic FUS causes early synaptic defects prior to aggregation in an ALS-FUS mouse model (3) and this will be the focus of this talk. Like TDP-43, FUS regulates nuclear RNAs, but its role at the synapse is poorly understood. After determining the sub-synaptic localization of FUS using super-resolution imaging, we employed CLIP-seq on synaptoneurosomes to identify synaptic FUS RNA targets. The latter encode proteins associated with synapse organization and plasticity. Significant increase of synaptic FUS during early disease in a knock-in mouse model of ALS-FUS was accompanied by alterations in density and size of GABAergic synapses. RNAs abnormally accumulated at the synapses of 6-month-old ALS-FUS mice were enriched for FUS targets and correlated with those depicting increased short-term RNA stability via binding primarily on multiple exonic sites. Our work indicates that early synaptopathy triggered by synaptic FUS accumulation, prior to aggregation, leads to ALS-FUS and understanding the underlying molecular events will be key for devising early and effective therapeutic interventions.

1. Afroz et al...Polymenidou (2017) Nature Communications, DOI: 10.1038/s41467-017-00062-0

2. Laferriere et al...Polymenidou (2018) Nature Neuroscience, DOI: 10.1038/s41593-018-0294-y

3. Sahadevan et al...Polymenidou (2020) bioRxiv. DOI: 10.1101/2020.06.10.136010

Aims

To determine whether cross-sectional and longitudinal cognitive and functional decline are associated with spread of transactive response DNA-binding protein of 43kDa (TDP-43) in the brain.

Methods

Longitudinal clinical-neuropathologic autopsy cohort study of 385 initially cognitively normal/mildly impaired older adults prospectively followed until death. Associations between TDP-43, amyloid-beta, tau neurofibrillary tangles (NFT), age, sex, genetics and clinical and neuropsychological scores and rates of their decline were investigated.

Results

Of 385 participants, 260 (68%) had no TDP-43 depoistion, 32(8%) had TDP-43 limited to amygdala, and 93(24%) had TDP-43 in the hippocampus and beyond. TDP-43-positive patients had higher frequency of APOE ε4 allele, neuritic plaque score and Braak NFT stage compared to TDP-43 negative patients; they were also less likely to remain cognitively normal at last examination, p<0.0001. Higher TDP-43 and Braak NFT stages independently were associated with faster decline in global cognition, functional performance, naming and episodic memory, whereas older age was associated with slower rates of cognitive, and functional decline. Cross-sectionally, higher TDP-43 and Braak NFT stages were associated with worse performance in the same measures as longitudinally observed (Figure 1). We also found higher amyloid-beta burden to be associated with worse global cognition and a higher frequency of behavioral changes. There were no associations with the APOE ε4 allele (Figure 1).

Conclusions

The associations between TDP-43 and faster rates of cognitive and functional performance over time suggests that TDP-43 is likely playing a role in the clinical progression to dementia, as is the case for tau.

Aims

To elucidate the genetic architecture of limbic-predominant age-related TDP-43 encephalopathy neuropathological change (LATE-NC), a common TDP-43 proteinopathy that is frequently comorbid with Alzheimer’s disease (AD) and increases dementia risk.

Methods

We studied the Religious Orders Study and the Rush Memory and Aging Project participants of European ancestry with imputed genotypes and post-mortem TDP-43 immunohistochemistry (n=1,180). TDP-43, amyloid-β (Aβ), and paired helical filament tau (PHFtau) were quantified using immunohistochemistry. LATE-NC burden was defined as a semi-quantitative six-point scale of TDP-43 burden averaged across six brain regions including amygdala, hippocampus, and neocortex. Genome-wide association study (GWAS) was performed using linear additive models controlling for age, sex, genotyping platform, and first three principal components from the genotype covariance matrix.

Results

Two distinct loci reached genome-wide significance in their association with higher LATE-NC burden: rs429358^C, an allele defining the APOE ε4 haplotype (beta=0.46, p=1.2×10^-12); rs17608393^A within GRM8 (beta=0.68, p=4.5×10^-8). There were 10 additional loci that reached pre-defined suggestive significance (p<10^-6). A suggestive variant rs72701368^G (beta=0.54, p=1.1×10^-7) tagged a large haplotype spanning the interferon-alpha (IFNA) gene cluster, and another suggestive variant rs56902449^A (beta=0.72, p=8.0×10^-7) was a cis-eQTL increasing the expression of MX2, a gene encoding interferon-induced GTP-binding protein Mx2. After adjusting for Aβ and PHFtau, the association between APOE rs429358^C and LATE-NC was partially attenuated (beta=0.28, p=2.6×10^-5), while GRM8 and IFNA variants’ association with LATE-NC remained similar.

Conclusions

AD-related APOE locus and AD-independent GRM8 locus were associated with LATE-NC burden, and converging evidence also implicated type I interferon pathway that is regulated by TBK1 (a FTD/ALS risk gene).

Aims

Intracellular aggregation of TDP-43 is found in most patients with amyotrophic lateral sclerosis (ALS), 45% of patients with frontotemporal dementia (FTD) as well as patients with limbic-predominant age-related TDP-43 encephalopathy (LATE). Having better tools to aid in more accurate and earlier diagnosis would substantially improve patient care. Thus, our aim is to provide the first tracer for the detection of TDP-43 aggregates by positron emission tomography (PET) to enhance diagnosis, staging, longitudinal measurement of disease progression and assessment of therapeutic efficacy in clinical trials.

Methods

Various methods including radiobinding and autoradiography were established to screen small molecules from our proprietary Morphomer^TM library. Both recombinant and TDP-43 aggregates obtained from brain tissue of patients with TDP-43 proteinopathies were used. For selected compounds, the physico-chemical properties, ADME and pharmacokinetic profiles were evaluated.

Results

Screening of the Morphomer^TM library led to identifying compounds binding to recombinant TDP-43 aggregates with Ki < 20 nM and confirmed on pathological TDP-43 derived from FTLD-TDP samples. Selectivity over other aggregation-prone proteins (amyloid beta, alpha-synuclein and Tau) was established. Target engagement on FTLD-TDP brain sections was demonstrated for selected compounds by autoradiography. Compounds with favorable CNS properties were profiled in pharmacokinetic studies.

Conclusions

Medicinal chemistry compound optimization and iterative design allowed the identification of compounds that bind to pathological TDP-43 with low nanomolar affinity. Compounds displaying suitable CNS pharmacokinetic profiles of a quick uptake with a fast and complete washout are being investigated to select a PET tracer for further development.

Aims

Cytoplasmic TDP-43 (TAR DNA-binding protein 43) inclusions are the pathological hallmark in most cases of amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). They are found not only in neurons, but also in astrocytes, which supply neurons with nutrients. Neuronal metabolism largely depends on the activation of astroglial adrenergic receptors, the primary target of noradrenaline, which in astrocytes triggers Ca²⁺ and cAMP signalling and augments aerobic glycolysis and lactate production. Cytoplasmic TDP-43 inclusions in astrocytes alone can cause motor neuron death, however, it is unclear whether this affects astroglial metabolism and ultimately the capacity of astrocytes to metabolically support neurons.

Methods

By using fluorescent dyes, genetically encoded FRET nanosensors and real-time confocal microscopy we measured the dynamics of Ca²⁺/cAMP signalling and lipid droplet (LD) and glucose metabolisms in isolated cortical astrocytes expressing the inclusion-forming C-terminal fragment of TDP-43 or wild-type TDP-43.

Results

The accumulation of LDs was increased in astrocytes with TDP-43 inclusions vs. astrocytes expressing wild-type TDP-43. These cells also exhibited reduced noradrenaline-mediated Ca²⁺ and cAMP signalling, likely due to the downregulation of β₂-adrenergic receptors. Although noradrenaline-triggered increase in intracellular lactate was similar in astrocytes with and without TDP-43 inclusions, the probability of activating aerobic glycolysis was facilitated in astrocytes with TDP-43 inclusions, while lactate MCT1 transporters were downregulated.

Conclusions

Our results show that, while noradrenergic signalling is reduced in astrocytes with TDP-43 inclusions, aerobic glycolysis and LD accumulation are facilitated, suggesting dysregulated metabolism that may affect astroglial metabolic support of neurons in ALS and FTD.

Aims

Accumulation of TDP-43 into intracellular inclusions is the hallmark of frontotemporal lobar degeneration-TDP (FTLD-TDP), amyotrophic lateral sclerosis (ALS) and limbic-predominant age-related TDP-43 encephalopathy (LATE) and present as co-pathologies in other neurodegenerative diseases. However, no therapeutic interventions targeting TDP-43 pathology are available. Antibody-mediated clearance of misfolded TDP-43 by microglia and inhibition of cell-to-cell protein spreading represents an attractive strategy for therapeutic intervention.

Methods

Monoclonal antibodies (mAbs) were generated against various regions of TDP-43 using our proprietary SupraAntigen^TM platform and selected for evaluation in cell-based and transgenic Tg(rNLS8) models of TDP-43 proteinopathies.

Results

High-affinity mAbs targeting different regions of TDP-43 displayed conformational selectivity to misfolded TDP-43 or all TDP-43 isoforms. One mAb, ACI-5891, binding to the C-terminal domain of TDP-43, demonstrated inhibition in vitro of de novo and seeded TDP-43 aggregation using recombinant and FTD brain-derived TDP-43 extracts, respectively. Furthermore, using mouse primary microglial cultures, ACI-5891 efficiently increased the capacity for cellular uptake of TDP-43 aggregates. When tested in the Tg(rNLS8) mouse model of ALS/FTLD-TDP, ACI-5891 significantly reduced the levels of phosphorylated TDP-43 and insoluble TDP-43 in the brain with a concomitant increase in the size of hypertrophic microglia.

Conclusions

From a panel of high-affinity, conformation-specific and pan antibodies, ACI-5891 was identified with unique properties for effectively inhibiting TDP-43 seeding and spreading in cellular assays and efficiently promoting microglia activity in vitro and in vivo. These characteristics demonstrated for the first time that an antibody targeting TDP-43 ameliorates TDP-43-mediated pathology in vivo providing validation for further development to target TDP-43-mediated neuropathology.

Abstract Body

As a central pathological hallmark, nuclear clearance of TDP-43 accompanied by its cytoplasmic aggregates in neurons and glia have been documented in ALS-FTD and Alzheimer’s disease (AD) patients with TDP-43 pathology. As recent studies in human disease support the view that loss of TDP-43 splicing repression underlies neurodegeneration, identification of TDP-43 loss-of-function as antemortem biomarkers will be critical for clinical application.

Since incorporation of TDP-43 nonconserved cryptic exons can be seen in brains of ALS-FTD and some AD cases, it may be possible that such loss of splicing repression can be detected in biofluids of patients with TDP-43 pathology. While some nonconserved cryptic exons can be fused in frame, giving rise to putative novel epitopes within the cryptic exon-encoded peptides, it would be possible to develop specific monoclonal antisera recognizing these neo-epitopes.

We have identified a set of human nonconserved cryptic exon targets for development of a panel of antibody probes to screen biofluids from patients. We have generated and characterized a battery of monoclonal antibodies recognizing nonconserved cryptic exon-encoded neopeptides.

Identification of these antigens in biofluids would not only serve as ante-mortem biomarkers to facilitate recruitment of patients as early as possible for clinical trials, but also to demonstrate target engagement and monitor efficacy of therapeutic strategy.

Current studies are designed to validate these monoclonal antisera for detection of nonconserved cryptic exons in CSF or blood of ALS-FTD as well as AD patients exhibiting TDP-43 pathology.