Beyond Alpha-Synuclein: Measuring Neuronal Function to Stratify Parkinson’s Patients
Parkinson’s drug development has more targets than ever. Programs now aim at the lysosome, the mitochondrion, the proteasome, LRRK2 kinase signaling and neuronal survival pathways, alongside the alpha-synuclein strategies that dominated the last decade. That breadth creates a measurement problem. If two patients carry the same synuclein pathology but reached it through different cellular failures, a trial has to tell them apart, enroll the right one, and show that the drug engaged the mechanism it was designed to correct. This article sets out why current biomarkers cannot do that, and how cell-of-origin isolation of neuron-derived extracellular vesicles from blood makes neuronal function measurable in living patients.
Is alpha-synuclein enough to stratify Parkinson’s patients?
Alpha-synuclein biomarkers confirm that Parkinson’s pathology is present, but they cannot show which cellular failure caused it. A Parkinson’s disease patient stratification biomarker also needs to identify the failing process in each patient, such as mitochondrial respiration, lysosomal degradation or kinase signaling, so trials can enroll patients whose biology matches the drug.
Protein biomarkers for alpha-synuclein, tau, TDP-43 and amyloid-beta remain the foundation of precision neurology. They answer whether pathology is present and how much of it there is. Stratification asks a different question: among patients who share a pathology, which upstream process put them there? Answering it requires a measurement that reports on what neurons are doing, not only on what they have accumulated.
One pathology, several upstream failures
The 2024 biological frameworks moved Parkinson’s from a clinical syndrome toward a biological entity. The neuronal alpha-synuclein disease integrated staging system (NSD-ISS) anchors diagnosis on pathological synuclein (S) and dopaminergic dysfunction (D) [1]. The SynNeurGe criteria add imaging evidence of neurodegeneration (N) and pathogenic gene variants (G) [2]. Both define who has the disease. Neither captures why.
The divergence is already visible. In the Parkinson’s Progression Markers Initiative (PPMI), the CSF alpha-synuclein seed amplification assay detected Parkinson’s with 87.7% sensitivity, but positivity fell to 67.5% in LRRK2 Parkinson’s and to 34.7% in LRRK2 carriers with normal olfaction [3]. And because more than 95% of people with Parkinson’s have idiopathic disease [4], the G axis leaves most patients without a mechanistic label. Whether alpha-synuclein is a driver, a propagating agent or a consequence may differ from one patient to the next.
Postmortem tissue shows what those differences look like.
- Mitochondrial. Stratifying idiopathic Parkinson’s by neuronal complex I deficiency identified a subtype covering roughly a quarter of cases, with widespread neuronal complex I loss, more mitochondrial DNA deletions and a tendency toward non-tremor-dominant phenotypes. The remaining patients showed no mitochondrial impairment outside the substantia nigra [4].
- Lysosomal. Glucocerebrosidase (GCase) activity in the substantia nigra fell by 58% in GBA mutation carriers and by 33% in sporadic patients [5], so lysosome-directed therapies may have responders well beyond the genetic subgroup.
- Cross-talk. Wild-type LRRK2 kinase activity is elevated in substantia nigra dopamine neurons in idiopathic Parkinson’s through an oxidative mechanism, with downstream defects in both mitochondrial protein import and lysosomal function [6].
Stratification will likely produce dominant-driver profiles rather than clean categories. A trial testing a mitochondrial agent still needs to know which patients have meaningful mitochondrial impairment in their neurons.
Why current biomarkers cannot sort patients by mechanism
In 2022, two Phase 2 trials of monoclonal antibodies against aggregated alpha-synuclein enrolled 316 and 357 participants with early Parkinson’s. Over 52 weeks, neither separated from placebo on the MDS-UPDRS or dopamine transporter imaging [7], [8]. Both enrolled by clinical criteria. Target choice, epitope and dose are all possible explanations. Biological heterogeneity is one the field cannot rule out without biomarkers that report on mechanism.
Closing that gap requires a blood-based measurement that meets three conditions.
1. It has to come from neurons. In PPMI, GCase activity measured in dried blood spots from 1,559 samples correlated with white blood cell count (r = 0.45) [9]. A pathway readout in whole blood reports leukocyte biology at least as much as neuronal lysosomes. For blood-based CNS biomarkers, knowing a signal’s tissue of origin matters more than how sensitively it is detected.
2. It has to report on the inside of the cell. CSF biomarker limitations are partly practical, since lumbar puncture burden caps how many timepoints a trial can collect. They are also biological. A free protein in CSF says little about whether a neuron’s respiratory chain or lysosomal machinery is working.
3. It has to work on routine samples, repeatedly. Mechanism-matched enrollment needs a baseline readout on every screened patient. Pharmacodynamic monitoring needs one at every visit. Validation needs one on biobanked plasma from cohorts that have already been followed for years. A method that depends on fresh samples, large volumes or specialist collection at every site fails all three uses before the biology is even tested.
How cell-of-origin NDEV isolation makes neuronal function measurable
Extracellular vesicles (EVs) are lipid-bilayer particles that cells release into biofluids, carrying proteins, lipids and RNA packaged inside the cell of origin. Neuron-derived extracellular vesicles (NDEVs) are a small fraction of plasma EVs, so the value of any NDEV biomarker depends on how specifically and reproducibly they are isolated. ExoSORT™ was built to meet the three conditions above.
Neuronal provenance, demonstrated rather than assumed
The workflow captures NDEVs with antibodies against two neuron-specific surface proteins, rather than a single marker with known expression outside the nervous system. In the method’s first peer-reviewed description, isolates contained 35-fold higher neuronal marker levels than an isotype-control pulldown. Hemoglobin, platelet factor 4 and albumin mRNA were depleted relative to plasma, while neuronal neurogranin mRNA was retained [10]. An independent validation led by Columbia University, with Johns Hopkins and Harvard T.H. Chan collaborators, found ~237-fold enrichment of synaptophysin in the neuronal fraction compared with isotype control and ~482-fold compared with total plasma EVs [11]. Both studies characterized isolates against MISEV guidelines, the International Society for Extracellular Vesicles’ reporting standard [12].
Neurons, astrocytes and microglia from the same plasma
The same immunoaffinity workflow enriched astrocyte-derived EVs, validated by GFAP, and microglia-derived EVs, validated by P2Y12 and a 23-fold enrichment of AIF1 over isotype control [11]. For Parkinson’s stratification this matters. A mitochondrial or lysosomal signal can be assigned to neurons rather than glia, and neuroinflammatory activity can be measured in its own compartment instead of contaminating the neuronal readout.
Intracellular cargo that reports on function
Because the vesicles are captured intact and then lysed, the readout is their internal cargo: proteins, mRNA and lipids, including the brain-enriched lipid 24-hydroxycholesterol [10]. Readouts run on NDEV isolates include synaptic proteins, proBDNF, and stress-response and inflammatory pathway proteins such as NRF2 and NF-κB. Synaptic proteins and proBDNF in these isolates already differ between early Alzheimer’s patients and controls [10]. The organelle markers this article argues for are measurable in neuronal EVs. Lysosomal proteins in neurally derived exosomes distinguished Alzheimer’s patients up to 10 years before diagnosis [13]. Complex I and complex III subunits were lower in putative neuron-derived exosomes from 28 Parkinson’s patients than from 33 controls [14]. In that exploratory study the combined markers reached only 51.5% specificity as a diagnostic, which is what the tissue data predict. If only about a quarter of idiopathic patients carry widespread neuronal complex I deficiency [4], a mitochondrial marker should split patients from other patients, not all patients from controls.
Precision and sample requirements that fit a trial
Repeated isolations of the same plasma samples across five independent experiments gave coefficients of variation of 8.0% to 22.7%. Variability between trained operators was lower than variability between donors, and the method transferred to an independent laboratory. Recovery of spiked neuronal EVs varied by less than 13% across plasma from healthy and dementia donors [10]. The workflow runs on small volumes of frozen plasma [11], which opens existing biobanks, not just new prospective collections, to mechanistic analysis.
The limitations are worth stating plainly. Any immunoaffinity method enriches antigen-positive EV subpopulations rather than every vesicle a tissue releases [11]. Absolute analyte levels can shift with preanalytical handling across cohorts [10], which makes standardized collection and internal reference samples part of the method rather than an afterthought. Functional NDEV markers are not yet qualified trial endpoints, and the Parkinson’s subtyping studies are still to be done.
Putting functional NDEV biomarkers to work in a Parkinson’s program
Parkinson’s work on this platform has started with synuclein itself. NeuroDex’s Michael J. Fox Foundation-funded program found differences in NDEV-associated alpha-synuclein between healthy volunteers and people with Parkinson’s or multiple system atrophy, and is extending the analysis to two larger established clinical cohorts [15]. Adding functional cargo to the same isolates is what turns a synuclein readout into a stratification panel.
| Program decision | NDEV readout | What is established | What still needs to be shown |
|---|---|---|---|
| Retrospective subtyping of an existing cohort | Mitochondrial, lysosomal and stress-response proteins in neuronal EVs from biobanked plasma | Neuronal enrichment, precision and small-volume frozen plasma workflow [10], [11] | Distributions that split Parkinson’s patients into subgroups aligned with phenotype or progression |
| Biological validation | Antemortem NDEV mitochondrial or lysosomal signal | Tissue-level subtypes exist [4], [5] | Concordance with neuronal complex I or GCase status in the same brains |
| Mechanism-matched enrollment | Baseline functional panel alongside synuclein status | NDEV synuclein differences in Parkinson’s and MSA [15] | Prospective cut-offs that enrich for the targeted mechanism |
| Pharmacodynamic monitoring | Change in the targeted pathway in neuronal EVs at each visit | Precision suitable for longitudinal sampling [10] | Dose-dependent shifts under a mechanism-targeted therapy |
| Glial confound control | Parallel astrocyte and microglia EV fractions | Validated enrichment from the same plasma [11] | Reference ranges for neuroinflammatory signal in Parkinson’s cohorts |
Consider how this would play out for a lysosome-directed program. At screening, each candidate’s plasma yields a neuronal EV fraction and parallel glial fractions. The neuronal fraction is profiled for lysosomal proteins alongside synuclein status, and the glial fractions show whether an apparent lysosomal signal is neuronal or inflammatory in origin. Enrollment then enriches for patients whose neuronal lysosomal profile falls outside the reference range, regardless of GBA status, since the tissue data suggest many sporadic patients share the deficit [5]. During dosing, the same lysosomal readout is repeated at each visit. A shift toward the reference range in the treated arm, and not in placebo, is evidence that the drug reached and changed the pathway in neurons. A flat readout in a well-exposed arm is equally informative, because it separates a failure of target engagement from a failure of the hypothesis before a clinical endpoint is ever read. The same design applies to a mitochondrial program, with complex subunits and stress-response proteins in place of lysosomal markers.
The highest-value first step is retrospective. Natural history cohorts with stored plasma, ideally including participants with postmortem tissue, can answer the core question without a new trial: does an NDEV mitochondrial or lysosomal signal measured in life identify the same patients that tissue analysis classifies after death? A positive answer gives translational teams something no current Parkinson’s blood biomarker provides, a patient-level estimate of which mechanism a drug should target. It also gives clinical development teams an enrichment strategy and a pharmacodynamic readout from the same sample.
From measuring pathology to measuring neurons
The biological definition of Parkinson’s tells trial teams who has synuclein disease. The widening pipeline makes the next step urgent: telling them why. Every program aimed at the lysosome, the mitochondrion or kinase signaling assumes it can find the patients whose neurons depend on that pathway, and without functional biomarkers that assumption goes untested until a Phase 2 readout. At NeuroDex, we built our isolation platform so that assumption can be tested from a blood draw, in the cell type that matters, as often as a trial needs. The same logic applies beyond Parkinson’s. In ALS and frontotemporal dementia, TDP-43 raises the same driver-or-consequence question, and answering it will need the same kind of measurement.
References
- Simuni T, Chahine LM, Poston K, et al. A biological definition of neuronal α-synuclein disease: towards an integrated staging system for research. Lancet Neurol. 2024;23(2):178–190. https://doi.org/10.1016/S1474-4422(23)00405-2
- Höglinger GU, Adler CH, Berg D, et al. A biological classification of Parkinson’s disease: the SynNeurGe research diagnostic criteria. Lancet Neurol. 2024;23(2):191–204. https://doi.org/10.1016/S1474-4422(23)00404-0
- Siderowf A, Concha-Marambio L, Lafontant DE, et al. Assessment of heterogeneity among participants in the Parkinson’s Progression Markers Initiative cohort using α-synuclein seed amplification: a cross-sectional study. Lancet Neurol. 2023;22(5):407–417. https://pmc.ncbi.nlm.nih.gov/articles/PMC10627170/
- Flønes IH, Toker L, Sandnes DA, et al. Mitochondrial complex I deficiency stratifies idiopathic Parkinson’s disease. Nat Commun. 2024;15(1):3631. https://doi.org/10.1038/s41467-024-47867-4
- Gegg ME, Burke D, Heales SJ, et al. Glucocerebrosidase deficiency in substantia nigra of Parkinson disease brains. Ann Neurol. 2012;72(3):455–463. https://doi.org/10.1002/ana.23614
- Di Maio R, Hoffman EK, Rocha EM, et al. LRRK2 activation in idiopathic Parkinson’s disease. Sci Transl Med. 2018;10(451):eaar5429. https://doi.org/10.1126/scitranslmed.aar5429
- Pagano G, Taylor KI, Anzures-Cabrera J, et al. Trial of prasinezumab in early-stage Parkinson’s disease. N Engl J Med. 2022;387(5):421–432. https://doi.org/10.1056/NEJMoa2202867
- Lang AE, Siderowf AD, Macklin EA, et al. Trial of cinpanemab in early Parkinson’s disease. N Engl J Med. 2022;387(5):408–420. https://doi.org/10.1056/NEJMoa2203395
- Alcalay RN, Wolf P, Chiang MSR, et al. Longitudinal measurements of glucocerebrosidase activity in Parkinson’s patients. Ann Clin Transl Neurol. 2020;7(10):1816–1830. https://doi.org/10.1002/acn3.51164
- Eitan E, Thornton-Wells T, Elgart K, et al. Synaptic proteins in neuron-derived extracellular vesicles as biomarkers for Alzheimer’s disease: novel methodology and clinical proof of concept. Extracell Vesicles Circ Nucleic Acids. 2023;4(1):133–150. https://doi.org/10.20517/evcna.2023.13
- Pierri B, Jackson GL, Gololobova O, et al. Immunoaffinity-based protocol to enrich nervous system cell-, lung alveolar cell-, and hepatocyte-derived extracellular vesicles from human plasma. J Extracell Biol. 2026;5(7):e70171. https://doi.org/10.1002/jex2.70171
- Welsh JA, Goberdhan DCI, O’Driscoll L, et al. Minimal information for studies of extracellular vesicles (MISEV2023): from basic to advanced approaches. J Extracell Vesicles. 2024;13(2):e12404. https://doi.org/10.1002/jev2.12404
- Goetzl EJ, Boxer A, Schwartz JB, et al. Altered lysosomal proteins in neural-derived plasma exosomes in preclinical Alzheimer disease. Neurology. 2015;85(1):40–47. https://doi.org/10.1212/WNL.0000000000001702
- Yao Y, Li Y, Zhao C, et al. Mitochondrial proteins in putative neuron-derived plasma exosomes are altered in Parkinson’s disease: an exploratory study. Mol Neurobiol. 2026;63(1):525. https://doi.org/10.1007/s12035-026-05817-9
- Michael J. Fox Foundation for Parkinson’s Research. Blood test for stratifying people with Parkinson’s disease based on extracellular vesicles derived from neurons. Funded grant summary. https://www.michaeljfox.org/grant/blood-test-stratifying-people-parkinsons-disease-based-extracellular-vesicles-derived-neurons
Leave a Reply