2025 Theses Doctoral

Astrocyte-Derived Extracellular Vesicle Based Biomarkers of Disease and Central Nervous System (CNS) Metal Levels in Amyotrophic Lateral Sclerosis (ALS)

Butt, Tanya H.

Background: There is an unmet need for minimally invasive biomarkers that directly reflect central nervous system (CNS) pathology and potential environmental contributors to disease in amyotrophic lateral sclerosis (ALS). Astrocyte-derived extracellular vesicles (ADEVs) offer a compelling solution because of their ability to cross the blood-brain barrier, carry astrocyte-specific cargo, and participate in processes implicated in ALS pathophysiology, including metal dyshomeostasis and TAR DNA-binding protein 43 (TDP-43) proteinopathy. However, neither the diagnostic potential of ADEV-based TDP-43 biomarkers nor the ability of ADEV metal content to predict CNS metal levels has been evaluated in ALS populations. In Chapter 1, we introduce the longstanding challenges in understanding the environmental and occupational etiology of ALS, the barriers to early diagnosis, and the current state of biomarker development in the field.

Objective: The objective of this dissertation is to determine whether ADEVs can serve as minimally invasive, CNS-specific biomarkers in ALS by: (1) evaluating the diagnostic utility of whole blood and plasma isolated ADEV TDP-43 and phosphorylated TDP-43 (pTDP-43) in predicting ALS case status and (2) assessing whether metals carried in ADEVs accurately reflect CNS metal levels, thereby providing a new tool for investigating the role of metal exposure and dyshomeostasis in ALS.

Methods: We developed a rigorously optimized and reproducible protocol for the isolation of ADEVs from plasma and whole blood. ADEV quality and enrichment was validated by measuring EV-, astrocyte-, and neuronal-specific protein markers, including the absence of a non-EV protein. Transmission electron microscopy and nanoparticle tracking analysis confirmed expected EV morphology and particle concentration. Tetraspanin cluster of differentiation 81 (CD81) was identified as an optimal normalization factor to account for ADEV isolation efficiency and the ratio of ADEV biomarkers to CD81 was consequently used for prediction modeling.

In Chapter 2, this protocol was applied to samples from three well-characterized ALS cohorts, the National ALS Biorepository, the ALS Multicenter Cohort Study of Oxidative Stress (COSMOS), and the Northeast Amyotrophic Lateral Sclerosis Consortium (NEALS), and matched controls. Diagnostic accuracy of ADEV TDP-43 and phosphorylated TDP-43 (pTDP-43) was evaluated by logistic regression and random forest models with cross-validation, adjusting for age and sex, and including 100 ‘seed’ randomization.

In Chapter 3, we adapted our ADEV isolation protocol for metal analysis and quantified metals in paired plasma, ADEVs, motor cortex, and lumbar spinal cord samples from veterans with ALS, a population disproportionately exposed to metals. We applied generalized additive models with penalized splines to determine whether metals measured in ADEVs or plasma best predicted CNS metal concentrations. Permutation testing assessed whether partial R2 estimates were due to chance, and bootstrap distributions of partial R2 estimates were compared between ADEV and plasma models. Because the interval between plasma and CNS sample collection ranged from 1 to 20 years, analyses were performed in all participants and in a subset with ≤ 4 years between plasma and CNS tissue collection.

Results: Plasma ADEV pTDP-43 CD81 ratio best predicted sporadic ALS case status with mean AUCs of 0.89 (95% CI: 0.75-0.99) and 0.86 (95% CI: 0.82-0.90) from logistic and random forest models, respectively. ADEV manganese and strontium more accurately predicted their corresponding motor cortex metal concentrations than plasma, and ADEV copper better predicted lumbar spinal cord copper than plasma. In the subset, ADEV manganese explained 42% and ADEV strontium explained 59% of their respective variation in motor cortex, while ADEV copper explained 20% of its variation in lumbar spinal cord.

Conclusion: These findings establish ADEVs as minimally invasive and CNS-relevant biomarkers capable of capturing TDP-43-related proteinopathy and CNS metal levels in ALS. Plasma-derived ADEV pTDP-43 demonstrates strong discriminatory ability for ALS cases, and metals in ADEVs reliably reflect manganese, strontium, and copper concentrations in the motor cortex and spinal cords of veterans with ALS. This work directly addresses translational gaps by establishing robust isolation methods for ADEVs and by evaluating ADEV TDP-43 pathology and metal content as novel predictors of ALS diagnosis and as proxies for CNS metal levels. By bridging environmental toxicology, epidemiology, neuroscience, and applied statistics, this dissertation advances a scalable platform for biomarker discovery that may improve disease detection, stratify mechanistic subtypes, and enhance epidemiologic investigations of metal-associated risk in ALS.