Profiling neuroinflammation in amyotrophic lateral sclerosis

BACKGROUND: Amyotrophic lateral sclerosis (ALS) is a neurodegenerative disease primarily characterised by progressive loss of motor neurons. ALS is highly heterogeneous and exists on a clinical and pathogenetic spectrum with frontotemporal dementia (FTD); those anywhere on this spectrum are said to have ALS-frontotemporal spectrum disorder (ALS-FTSD). Accordingly, for patients carrying mutations such as the C9orf72 hexanucleotide repeat expansion, it is difficult to predict aspects of disease presentation such as disease duration, or whether people will experience motor symptoms, cognitive symptoms, or a combination of both. Such disease heterogeneity can be a significant confounder in the meaningful measurement of outcomes in clinical trials. Thus, a better understanding of factors underlying disease heterogeneity, as well as treatment-relevant subgroups for appropriate trial stratification, is paramount. Previous studies have identified differences in inflammatory factors detected in cerebrospinal fluid between ALS and FTD patients, and recent clinical trials have reported differences in circulating inflammatory mediators between treatment responders and non-responders. This considered, this dissertation hypothesises that inflammation contributes to disease heterogeneity in ALS, and tests this hypothesis by comprehensively profiling neuroinflammation in ALS, particularly in C9-ALS-FTSD with validation in other ALS cohorts (i.e., sporadic ALS-FTSD, SOD1-ALS). METHODS: High-throughput immunohistochemistry, digital pathology analysis, and random forest modelling were employed to characterise glial activation and RNA-binding protein localisation and aggregation in motor and extramotor brain regions of deeply clinically phenotyped C9-ALS-FTSD post-mortem cases (n = 10), compared with age- and sex-matched controls (n = 10). Guided by the results of this analysis, C9-ALS-FTSD post-mortem motor cortex tissue was further profiled using a NanoString nCounter neuroinflammation panel, with findings validated in sporadic ALS-FTSD (n = 18) and SOD1-ALS (n = 5) tissue and an independent public dataset. Further interrogation of candidate pathways and their roles in heterogeneity and cell type-specific disease mechanisms was then conducted with a variety of techniques including immunohistochemistry, BaseScope™ in situ hybridisation, cell type specificity analysis, split-pool ligation-based transcriptome sequencing (SPLiT-seq), and stem cell modelling of C9-ALS microglia. RESULTS: Digital pathology analysis revealed significant microglial activation in C9-ALS-FTSD, a significant correlation between macrophage staining abundance and phosphorylated TDP-43 aggregate staining abundance, as well as increased macrophage activation in the language-related region BA39 in language-impaired cases. Random forest modelling identified that microglia/macrophage- and FUS-related staining were accurate classifiers of C9-ALS-FTSD versus control status. Further characterisation of neuroinflammatory dysregulation on a molecular level with NanoString revealed clinical correlates of disease duration (i.e., BDNF expression) and cognitive function (i.e., FKBP5 expression), increased glial NF-κB activation in C9-ALS-FTSD, and two distinct neuroinflammatory panel signatures (NPS1, NPS2) that also existed across sporadic ALS-FTSD and SOD1-ALS cases and that did not segregate clearly with known clinical or pathological features. A filtered NPS-defining gene list was obtained, highlighting the upregulation of pro-inflammatory genes and downregulation of synaptic signalling and axonal transport genes in NPS1 and vice versa in NPS2. Cell type specificity analysis of NPS-defining genes confirmed a role for glia and excitatory neurons in defining NPS. Exploratory iPSC-derived C9-ALS microglia experiments provided pilot data regarding the dependence of microglial activation and cytokine release on the NF-κB and BDNF signalling, as well as outlined a workflow for further characterisation of these pathways. CONCLUSIONS: The findings in this dissertation provide highly resolved insight into neuroinflammatory dysregulation and neuroinflammatory subgroups in ALS, demonstrating a potential link between macrophage activation, TDP-43 aggregation, and cognitive impairment. The dissertation also highlights BDNF and FKBP5/NFκB as candidate prognostic biomarkers, and additionally offers a robust starting list of candidate markers for NPS subgroup classification and detection in peripheral samples. Importantly, these signatures did not segregate with demographic (i.e., genetics, sex) or clinical (i.e., disease duration, region of onset) data, emphasising that these are not signatures that can be easily identified through clinical examination or classical stratification approaches. Thus, these subgroups underscore the need for molecular stratification to achieve effective clinical trial design and meaningful outcome measurement.

Grantor dc:publisher

The University of Edinburgh

Year dc:date.issued

2024

Author dc:creator

• Rifai, Olivia M.

Advisors dc:contributor.advisor

• Sibley, Christopher
• Abrahams, Sharon
• Chandran, Siddharthan

• amyotrophic lateral sclerosis
• ALS
• neuroinflammation
• neuropathology
• microglia

Language dc:language.iso

en

Identifier URI

http://dx.doi.org/10.7488/era/4065

OAI identifier oai:identifier

oai:era.ed.ac.uk:1842/41330