Comparative assessment of Alzheimer’s disease-related biomarkers in plasma and neuron-derived extracellular vesicles: a nested case-control study

Introduction: Alzheimer’s disease (AD) is currently defined according to biomarkers reflecting the core underlying neuropathological processes: Aβ deposition, Tau, and neurodegeneration (ATN). The soluble phase of plasma and plasma neuron-derived extracellular vesicles (NDEVs) are increasingly being investigated as sources of biomarkers. The aim of this study was to examine the comparative biomarker potential of these two biofluids, as well as the association between respective biomarkers. Methods: We retrospectively identified three distinct diagnostic groups of 44 individuals who provided samples at baseline and at a mean of 3.1 years later; 14 were cognitively unimpaired at baseline and remained so (NRM-NRM), 13 had amnestic MCI that progressed to AD dementia (MCI-DEM) and 17 had AD dementia at both timepoints (DEM-DEM). Plasma NDEVs were isolated by immunoaffinity capture targeting the neuronal markers L1CAM, GAP43, and NLGN3. In both plasma and NDEVs, we assessed ATN biomarkers (Aβ42, Aβ40, total Tau, P181-Tau) alongside several other exploratory markers. Results: The Aβ42/Aβ40 ratio in plasma and NDEVs was lower in MCI-DEM than NRM-NRM at baseline and its levels in NDEVs decreased over time in all three groups. Similarly, plasma and NDEV-associated Aβ42 was lower in MCI-DEM compared to NRM-NRM at baseline and its levels in plasma decreased over time in DEM-DEM. For NDEV-associated proBDNF, compared to NRM-NRM, its levels were lower in MCI-DEM and DEM-DEM at baseline, and they decreased over time in the latter group. No group differences were found for other exploratory markers. NDEV-associated Aβ42/Aβ40 ratio and proBDNF achieved the highest areas under the curve (AUCs) for discriminating between diagnostic groups, while proBDNF was positively associated with Mini-Mental State Examination (MMSE) score. No associations were found between the two biofluids for any assessed marker. Discussion: The soluble phase of plasma and plasma NDEVs demonstrate distinct biomarker profiles both at a single time point and longitudinally. The lack of association between plasma and NDEV measures indicates that the two types of biofluids demonstrate distinct biomarker signatures that may be attributable to being derived through different biological processes. NDEV-associated proBDNF may be a useful biomarker for AD diagnosis and monitoring.


Figure legends
Figure S1.Assessment of hemolysis using a previously described method (Harboe, 1959).
Series of gradually hemolyzed standards were prepared by diluting a completely hemolyzed sample (100%) with non-hemolyzed plasma (0%).(A) Absorbance of Hemoglobin at 415 nm in standards; (B) Hemoglobin absorbance of standards and 20 plasma samples isolated using the methodology described in the manuscript; (C) Visual representation of results.GluR2, and ENO2, that shows neuronal enrichment in NDEVs, compared to Total EVs, and beads only; (B) Validation of the isolation method, using a Human ProcartaPlex™ Simplex™ assay, that shows enrichment of canonical EV markers in NDEVs and pan-tetraspanin-EVs, compared to Total EVs, neat plasma, and beads only.TEV = Total EVs; BO = beads only; NeuN = Neuronal nuclear protein; GluR2 = Glutamate ionotropic receptor AMPA type subunit 2; ENO2 = Enolase 2; panTET-EVs = pan-tetraspanin-EVs.Values presented are mean (standard error).

Figure S5 .
Figure S5.Determination of optimal particle per frame (PPF) range using 200 nm beads.(A) PPF in function of sample fold dilution; (B) PPF range showing a linear relationship with sample dilution; (C) Size distribution of beads at different concentrations/PPF; (D) Assessment of the effect of sample concentration (PPF) on particle size mode.

Figure S6 .
Figure S6.Validation of optimal particle per frame (PPF) range using crude total EVs isolated from human plasma.(A, C, E) PPF in function of sample fold dilution; (B, D, F) PPF range showing linear relationship with sample dilution in crude EVs from the plasma of three different subjects;(G) Mean size distribution of crude EVs from the plasma of one human subject analyzed at concentrations providing PPFs either within the optimal range resulting in a linear relationship with sample dilution (red line), or at lower PPFs (black line).EVs/mL values have been corrected for dilution factor; (H) Min-to-max particle size mode distribution of crude EVs from the plasma of three human subjects analyzed at PPFs under (low PPF), within (optimal PPF), or above (high PPF) the optimal range.Statistical analysis: two-way ANOVA.

Table S1 .
Summary of non-Box-Cox transformed marker measurements