Single step capture and assessment of multiple plasma extracellular vesicle biomarkers in Alzheimer's disease detection

Abstract

Background

Blood tests for Alzheimer's disease (AD) that measure biomarkers related to neuropathology have demonstrated to be useful, minimally-invasive ways to identify patients for screening into clinical trials. While some AD biomarkers can be detected in plasma, greater sensitivity is needed to make plasma AD tests more effective. Extracellular vesicles (EVs) in plasma carry AD-related biomarkers from the brain and could offer a concentrated source of brain-related biomarkers, though the methodological complexities involved in isolating plasma EVs have hampered its validation for clinical use.

Objective

To explore the feasibility and effectiveness of developing blood tests for AD utilizing extracellular vesicle-bound protein biomarkers.

Methods

We developed a simplified method for isolating EVs directly from plasma using an alternating current electrokinetic (ACE) microchip. No sample pretreatment steps were needed. Protein biomarkers on the EVs were detected by adding fluorescent antibodies to the plasma samples before capture by the chip. This allowed measurement of EV biomarker levels directly on the chip.

Results

AD or non-AD control plasma was measured for ten different AD-related biomarkers. EV-associated NCAM1, pTau231, α-synuclein, and TDP-43 levels were able to distinguish a group of 10 AD, 10 mild cognitive impairment (MCI), and 10 non-AD subjects. pTau231 was different between AD and non-AD (p = 0.0300) and α-synuclein differentiated AD from MCI (p = 0.0148).

Conclusions

This study shows how ACE microfluidic chip technology can help differentiate AD and MCI patients from non-AD controls with clinical relevance. This work also highlights the important diagnostic role of plasma EV biomarkers in neurodegenerative disease.

Keywords

Alzheimer’s disease amyloid biomarkers blood extracellular vesicles neural cell adhesion molecule L1 neurodegenerative diseases p-Tau proteins synucleins TDP-43

Introduction

Alzheimer's disease (AD) is the most common cause of dementia with no cure or established pathogenesis. Efforts to develop effective treatments and reduce morbidity have been hampered by the barriers to timely diagnosis: patients most often do not present clinically until symptoms are quite advanced, when neurological damage is already extensive and irreversible. Diagnosis typically relies on brain imaging, such as positron emission tomography (PET) to identify pathologic changes in brain structure, or on invasive cerebrospinal fluid (CSF) sampling to assess the biomarkers amyloid-β (Aβ) and phosphorylated Tau (pTau).^1–3 There is a great clinical need for earlier, affordable, and less invasive diagnostic and prognostic tests for broader applications.^2,4,5

Growing evidence supports the use of biomarkers for early detection of AD. Changes in brain morphology and accumulation of Aβ plaques have been shown to occur up to 30 years before development of cognitive symptoms, with the characteristic tau tangles appearing later.^2,3,6–11 Elevations in plasma pTau231 may be detected very early in the AD continuum, before increases in pTau181 are observed, and plasma pTau217 levels can best predict progression from mild cognitive impairment (MCI) to AD; use of these pTau biomarkers has been demonstrated to be useful for accelerating screening of amyloid PET positive, asymptomatic participants for disease-modifying trials.^3,9,12–14 Some success in detecting AD-related biomarker signals directly in plasma or serum has been achieved using sensitive immunoassays, such as pTau181, but more sensitive and comprehensive analyses are complicated by the low abundance of these biomarkers in the general circulation.^14–16

Recently, extracellular vesicles (EVs) have been shown to carry brain-related protein markers across the blood-brain barrier into circulating blood. EVs are a type of vesicle (∼30–1000 nm) actively produced by cells to perform roles that include communication among cells, tissues, and the extracellular matrix, and transfer of various cargo molecules.^17–19 Importantly, circulating EVs have been shown to mediate intercellular transfer of pathogenic forms of the biomarkers α-synuclein, Aβ, and pTau, and to contain other brain-related proteins such as L1CAM (L1 cell adhesion molecule).^20,21 Accumulating evidence suggests that assays for markers of brain pathology perform better when using purified EVs than for plasma.^{15,16,20,22–26} However, to assess biomarkers found on circulating EVs, most methods typically require plasma or serum samples to undergo several pre-processing steps before EVs can be purified and analyzed. These methods include ultracentrifugation, polymer precipitation, or immunoaffinity methods that isolate only selected EV populations. Thus, the multi-step processes entailed in AD biomarker analysis for EVs have to date limited their clinical utility.

We have previously shown that an alternating current electrokinetics (ACE)-based microchip platform can be used for concurrent capture and analysis of bulk EVs directly from blood or plasma, for streamlined analysis of cancer biomarkers.^26–28 The current study demonstrates the first use of an AC electrokinetic microchip system for sensitive and rapid detection of ten brain-related plasma EV markers associated with the presence of AD. The method can readily assess levels of multiple AD-associated proteins on circulating EVs, providing for differentiation of clinically diagnosed AD and MCI patient samples from normal donor controls.

Methods

Sample collection and processing

Individual plasma specimens for this retrospective study were obtained from the Alzheimer's Disease Research Center at the University of California, San Diego. All participant samples were consented for biobanking and research and all samples were de-identified. AD diagnosis was confirmed by postmortem neuropathological examination. All AD cases were selected to exclude known comorbidities, such as Parkinson's disease or limbic-predominant age-related TDP-43 encephalopathy (LATE). MCI was determined by clinical diagnosis: Mini-Mental State Examination (MMSE) score 23–27. Pooled AD+ and non-AD plasma samples were made using 5 individual samples obtained from biobanks (BioIVT, Westbury, NY; PrecisionMed, Carlsbad, CA). To obtain plasma, whole blood samples were collected in K₂EDTA plasma collection tubes and processed within 4 h, using two low-speed centrifugation steps at 1500 X g for 10 min each. All specimens were then frozen within 1 h and stored at −80 °C until use. All frozen samples were maintained by the individual biorepositories.

Cell culture purified control EVs

To serve as positive controls for antibody optimization, control EVs were purified from cell culture supernatants of U138MG or U87MG cells, which collectively express all ten of the biomarkers studied here (see https://www.proteinatlas.org). To serve as a general positive control for the microchip function, EVs were isolated from cell culture supernatants of H1975 lung carcinoma cells known to express CA19-9 (ATCC #HTB-16, #HTB-14, #CRL-5908). Cell culture supernatants were first clarified by centrifugation at 3000 RCF for 20 min, then subjected to ultracentrifugation at 100,000 RCF for 2 h. Following resuspension of the pellets in phosphate-buffered saline (PBS), EVs were stored at −80 °C, and are referred to as “UC-purified EVs”. For all purified cell culture EVs, particle concentrations were measured by Nanotracking Analysis (NTA). For control experiments, purified EVs were spiked into plasma samples at a specific range of pre-determined concentrations.

ELISA assays

To verify the presence of each biomarker on the purified cell culture EVs, both U87MG and U138MG-derived UC-purified EVs were tested in ELISA kits selective for each biomarker. Because most of the neuronal biomarkers targeted in this study are predicted to localize to the EV lumen, steps were taken to disrupt the EV membranes to make internal proteins accessible to analysis. Purified EVs were diluted into PBS at the relevant particle concentrations, and 2X lysis buffer (Ray Biotech #EL-lysis) was added according to manufacturer's directions. The exosome mixture was sonicated for 5 min in a bath sonicator, then vortexed gently. The sonication and vortex steps were repeated in order 2 more times, then the solubilized samples were analyzed by ELISA according to the manufacturer's directions.

Human ELISA kits designed for plasma or serum inputs were purchased from CUSABIO (amyloid-β 1-40, CSB-E08299h; amyloid-β 1-42, CSB-E10684h; L1CAM, CSB-E09145h; NCAM1, CSB-E015511HU; alpha-synuclein, CSB-E18033h; TDP-43, CSB-E17007h; and total Tau, CSB-E12011h), RayBiotech (pTau181, PEL-Tau-T181-Q-1), or TriBioscience (pTau231, #TBS3296). No commercial ELISA kit suitable for plasma was readily available for pTau396.

Antibodies

Primary antibodies were purchased from Invitrogen (amyloid-β-40, #44-348A, rabbit; amyloid-β-42, #44-344; pTau231, #MN1040, mouse; pTau396, #710298, rabbit), Abcam (L1CAM, AB238501, rabbit; alpha-synuclein, #AB138501, rabbit), Santa Cruz Biologicals (Total Tau, #SC32274, mouse; NCAM1, #SC106, mouse), Cell Signaling Technology (pTau181 #12885S) and Protein Tech (TDP-43 #10782-2-AP, rabbit). CD81 was used at 1:500 in western blots and was from Abcam (ab219209). Secondary antibodies were from Jackson Immunoresearch Laboratories (West Grove, PA, USA; Alexa Fluor 488-F(ab’)₂ goat anti-mouse IgG #115-546-14; Alexa Fluor 488- AffiniPure F(ab’)₂ fragment donkey anti-rabbit IgG).

ACE chip EV collection and labeling

To test patient plasma, samples were thawed before use, and 20 μL of each sample was added to 40 μL of 0.5 X PBS containing diluted primary and secondary antibodies plus saponin at 0.01% final concentration (Sigma Life Sciences # 47036-250G-F). Following incubation for 60 min at 25 °C, the sample mixture was pipetted into one loading well of an 8-chamber microfluidic cartridge enclosing the ACE microelectrode chip (Biological Dynamics). Tris-EDTA buffer (TE; Promega #V6232) was added to the wash loading well, and the cartridge inserted into an OmniVerita instrument (Biological Dynamics, Inc.) to energize, wash, and quantify the resulting fluorescent images as previously described.²⁶ The automatic run involves a 5-min chip conditioning step, followed by a 10-min capture step, a 10-min wash step, and lastly an image acquisition step (∼30 min total). The output images were quantified using Trace Image Quant proprietary software (Biological Dynamics). Each sample was run a minimum of 3 times.

Figure 1.

Schematic illustration of OmniVerita process for detecting EVs. One-step characterization of EV biomarkers related to Alzheimer's disease and other dementias starts with the drawing and preparation of plasma from whole blood. Subsequently, antibodies are added together with detergents to permeabilize EV membranes, allowing access to intra-lumenal proteins. Samples are loaded into the platform and the EVs are isolated onto the microelectrode array. The marker levels are then quantified using in situ fluorescent microscopy.

To collect purified EVs from the chip without the antibody labeling steps, as previously described, 240 μL of each sample was flowed across a single-chamber chip at 3 μL/min for 120 min.²⁶ Following a 30-min washing step, the electrical signal was turned off, and the released EVs were collected in 35 μL of buffer. Western blotting, NTA, Bioanalyzer runs, and cryo-transmission electron microscopy (TEM) were also performed as previously described, with one addition. To prepare lysed EV samples for western blotting, following mixing of the samples in gel loading buffer, the exosome mixture was sonicated for 5 min in a bath sonicator, then vortexed gently. The sonication and vortex steps were repeated in order 2 more times, then the solubilized samples were heated for 5 min at 95 °C. Without the lysis step, most of the highly bouyant EV material remains in the stacking gel and does not penetrate the separating gel.²⁹

For negative and positive chip run controls, 2 chambers of every chip contained CA19-9 antibody (LS Bio #LS-B16973) and the appropriate 2° antibody (Alexa Fluor 488-F(ab’)₂ goat anti-mouse IgG), with (positive control) or without (negative antibody control) H1975 cell UC-purified EVs, spiked into healthy control K₂EDTA plasma. This ensured proper performance for each chip.

Antibody dilution optimization

To find the optimal concentration for each antibody used, contrived samples were created using UC-purified EVs (verified by ELISA as described above for expression of each target protein; see Supplemental Table 1). EVs were spiked at a range of concentrations into a mixture containing 20 μL pooled K₂EDTA plasma from healthy donors, 40 μL of 0.5X PBS, and various concentrations of the primary antibody of interest plus the appropriate secondary antibody. As negative antibody controls, samples were made without the addition of the EVs containing the antigen. All samples were run in the OmniVerita instrument, and results quantified as described above.

Results

Single step isolation and characterization of EVs

The utility of AC electrokinetic microfluidics (ACE) for isolation of EVs, including exosomes, based on size has previously been demonstrated.^26–28 Unlike many other EV purification methods, the use of ACE forces to segregate particles from biological fluids does not require any pre-processing steps (Figure 1). All particles within the size range of about 30 nm to 500 nm are captured by the chip electrodes; it is important to note that no steps were taken to selectively isolate neuron-derived EVs (Figure 2). Any specificity is provided by the antibody labeling. During capture, the EVs can be labeled with antibodies targeting cell type- or disease-specific markers, and with detergent permeabilization, the range of markers extends to those within the EV lumen (cytosolic proteins). Notably, these internal proteins include most of the traditional AD markers. This leads to an efficient process for cohesive capture and quantification of these markers within less than 30 min.

Figure 2.

Characteristics and mechanics of the OmniVerita microfluidic chip. Biological fluids (with typically high conductivity) containing EVs are added to a fluid cell in contact with the microelectrode array chip surface. When subjected to alternating current electrokinetics (ACE), the charges in a particle will reverse accordingly, resulting in movement either towards or away from the electrode. The net movement depends upon particle size: when tuned to a particular voltage and frequency, particles of about 50–500 nm in diameter are captured by the electrodes. Larger or smaller particles, including whole cells, cell debris, and some viruses, remain in the low field regions and are removed with a buffer wash.

To investigate the nature of the material captured by the ACE chip, pooled healthy donor plasma were run using the purification protocol, in which the particles are eluted from the chip (Figure 3). Cryo-TEM experiments show the presence of vesicles of about 100–200 nm in diameter, consistent with the presence of EVs. Bioanalyzer visualization of the total protein present in ultracentrifuge-purified material shows the presence of several large protein peaks that are not present in the ACE preparations. NTA results for ACE-purified particles show a mean diameter of 101.0 ± 3.5 nm, and the presence of CD81 as seen by western blotting also suggests that the particles collected by the ACE chip are EVs.

Figure 3.

Characterization of particles captured by ACE chips from plasma. Pooled healthy donor plasma was run on ACE chips, and captured particles were eluted in buffer. (a) NTA measurements for 200 purifications demonstrate particle diameters, with a median diameter of 101.0 ± 3.5 nm. (b) Western blots show the presence of CD81 in EVs from cell cultured U138MG cells purified by UC, or purified from AD, MCI, or normal donor plasma using the ACE chips; whole healthy normal plasma shows the absence of detectable CD81. (c) Bioanalyzer runs show comparison of total protein content for ACE-purified (top) and ultracentrifuge-purified EVs. Arrows indicate protein peaks present in the ultracentrifuged EVs, but lower or not present in the ACE purified EVs. (d) ACE-purified material examined by cryo-transmission electron microscopy (cryo-TEM) shows the presence of cup-shaped vesicles, consistent with EVs.

When control UC-purified EVs are spiked into pooled normal plasma and run on the ACE chip system, the captured EVs can be labeled with the EV-selective markers CD63 and CD81; the specificity of the system is illustrated by the absence of signal when primary antibodies are omitted (Figure 4(a), top). The bottom panels of Figure 4(a) display the quantified results corresponding to the CD63- and CD81-labeled images shown above in the top panels. To verify that the system also works for clinical samples, endogenous CD63 is shown to be present in donor plasma EVs isolated onto the ACE chip electrodes (Figure 4(b), top). The brightfield image of the same field (bottom panel) shows the locations of the circular electrodes. For both CD63 and CD81 labeling, permeabilization is unnecessary due to the transmembrane localization of these proteins. However, most of the AD biomarkers investigated here typically localize to the EV lumen. To ensure efficacy of the microfluidic chip system for use with markers contained within the EV membranes, permeabilization with mild detergent was used to ensure access of the antibodies (Figure 4(c)).

Figure 4.

Detection of EV-associated marker proteins in material captured by the ACE chip. (a) Top: ultracentrifuge-purified H1975 cell culture EVs were captured onto ACE chip microelectrode array and tested for the presence of EV-associated proteins CD63 and CD81 (top images) using indirect immunofluorescence. As controls for nonspecific 2° antibody binding, experiments were repeated, but with the 1° antibody omitted (bottom images). Bottom: bar graphs depict quantification of fluorescence signals for CD63 (left) and CD81 (right). (b) Detection of CD63 using Alexa Fluor 488 fluorescent anti-CD63 antibodies (top image, viewed in FITC channel) in material collected onto the chip microelectrode array from ≤20 µL of patient plasma (bottom, same image viewed in brightfield). (c) Addition of detergent enhances access of antibodies to proteins found within the EV bilayer membranes (no detergent, top left images; with detergent, top right images).

AD biomarker detection

Ten markers across the AD spectrum were chosen for analysis (Supplemental Table 1). This set included both traditional AD markers, such as Tau and selected pTau species, Aβ variants, and other emerging markers that to date have been difficult to quantify such as pTau396, TDP-43, or α-synuclein. Also included were L1CAM and NCAM1, markers for neuron-derived EVs. Antibodies directed against each of these targets were sourced commercially. To test and optimize each antibody for use with the intact EVs collected onto ACE chips, it was necessary to devise a model system of contrived positive controls using EVs. To this end, UC-purified EVs were obtained from two cultured glioblastoma cell lines (U87MG and U138MG), which collectively express all ten markers examined here (proteinatlas.org). Biomarker expression in the EV preparations was experimentally verified by commercial ELISA assays (Supplemental Table 1). Because most of these proteins reside within the lipid bilayer membrane of the EVs, and the EVs have low buoyant density, a method was developed to solubilize and fragment the EVs for analysis by ELISA. Results are shown for nine markers; for pTau396, a commercial ELISA kit suitable for plasma and serum was not currently available.

When run on the ACE chip system, each of the ten antibodies showed a dose-response relationship with the model system positive control EVs (Figure 5(a)). For eight of the targets, U138MG cell EVs captured onto the electrodes showed the characteristic saturation curves; for the amyloid targets, U87MG EVs worked better to demonstrate the dose-dependent response. Interestingly, each target marker/antibody pair appears to generate a characteristic curve, even where the EVs used were the same. Representative images show raw data for the capture of one analyte, TDP-43, using 4 different concentrations of U138MG EVs (Figure 5(b)).

Figure 5.

Model system demonstrates detection of neuronal EV proteins on ACE microchips. (a) Use of model system to demonstrate dose-response detection of ten different AD biomarkers by selected antibodies. EVs purified by ultracentrifuge from U138MG or U87MG cultured cells were added to pooled healthy donor plasma at a range of concentrations, then run in the OmniVerita system. Antibodies selective for each AD biomarker protein are shown separately. (b) Images show representative results for detection of TDP-43 in EVs collected onto the OmniVerita chip. Quantification of pixels in each single image produced one data point for curves shown in Part A. Values represent amounts of TDP-43 protein contained in the U138MG EVs that were loaded into each sample well of the chip.

Performance of optimized assays with clinical samples

To ensure that successful detection of each analyte using the spiked EV model samples could translate to detection in clinical samples, a pool of AD+ patient plasma (N = 5) was tested using the ACE chip system. When compared to a pool of control non-AD plasma (N = 5), results show clear differentiation of AD+ plasma for all ten analytes (Figure 6(a) and (b)). Signals ranged from 2.3X the control levels, for NCAM1, to as much as 63.9X control levels for pTau181. It is important to note that differences in individual antibody affinities cannot be distinguished here from differences in actual biomarker expression on individual EVs. Overall, these results point to the potential use of the ACE chip system to detect differences in AD-related biomarkers in patient plasma EVs.

Figure 6.

Performance of optimized antibodies with pooled AD or control plasma. (a) Differentiation of pooled AD patient plasma (top images) from pooled healthy donor plasma (bottom images) is shown for each of ten AD biomarkers using EVs isolated by the OmniVerita chip. (b) Quantification of results for each marker in A above shows clear discrimination of AD patient samples from healthy donor samples. All plasma samples were collected in K₂EDTA tubes.

To further investigate the clinical potential of the ACE chip assay using selected EV biomarker proteins to stratify patient samples, a small case-control cohort of patient plasma was obtained with clinical diagnoses of AD, MCI, or non-AD (Figure 7(a) and Supplemental Table 2). For these samples we focused analysis on NCAM, as a marker of neuronal EVs, and on three other biomarkers for which performance in plasma is poor and where there are few reliable testing options currently available. Results for NCAM1, pTau231, α-synuclein, and TDP-43 are shown in Figure 7(b) and demonstrate the ability of the ACE chip system to detect and compare levels of these protein markers in total plasma EVs. NCAM1, which may reflect the level of neuronal-related EVs in circulation, was not significantly different among the patient cohorts but was seen to vary widely among individuals. Two markers, pTau231 and α-synuclein, were found to differentiate AD from either healthy control (HC) or MCI cohorts (p = 0.030 and p = 0.0148, respectively; Figure 7(b)). While TDP-43 did not show significant differences between each cohort for this particular sample set, which did not include potential comorbidities such as TDP-43 encephalopathy (LATE), the marker did show the capability to stratify the populations, in particular the “healthy” cohort. Multivariate analysis for two biomarkers, TDP-43 and pTau231, illustrates greater cohort differentiation, with the HC group clustered into one region of the graph (Figure 7(c)).

Figure 7.

Plasma EV biomarker levels in AD, MCI, or control samples. (a) Summary of small cohort demographics. (b) Plasma EV levels of four different AD biomarkers in AD+ and MCI patients or healthy controls. P-values are shown where significant. Boxes show interquartile range; horizontal lines show means. (c) Multivariate graph shows TDP-43 and pTau231 ACE assay results for AD, MCI, and HC cohorts; TDP-43 results are inverted left-right for clarity. Arrows indicate HC samples with high levels of both amyloid-B42 and 40, but normal ratios.

To compare the sensitivity of patient sample OmniVerita results for EVs with measurements made on commercial ELISA kits, standard curves were generated using kits specific for NCAM1, α-synuclein, pTau231, and TDP-43 (Supplemental Figure 1). Each biomarker was measured for purified U138MG EVs using both the ELISA and OmniVerita systems, and results were compared. This provided a way to convert the Omniverita AFU measurements to a concentration provided by the ELISA, using a conversion factor. Patient sample OmniVerita results were plotted onto the ELISA standard curve plots to show a direct comparison of signal ranges. For three of the biomarkers, the OmniVerita assay is much more sensitive. EV levels of α-synuclein in patient plasma are clustered in the lower range of detection, and for pTau231 and TDP-43, EV levels are below the range of detection. This demonstrates that measurement of blood-based EV biomarker levels may be an important factor in early AD detection, but sensitivities needed for EVs may require new types of assays. The OmniVerita assay provides a way to differentiate AD and MCI from healthy states using blood-based EV markers.

Discussion

In this study, we demonstrated feasibility of a rapid plasma EV biomarker test to detect the presence of AD or MCI. We showed that EVs captured from peripheral blood contain multiple markers of neurodegenerative disease, and that use of the ACE-driven microfluidic microelectrode array chip can provide simultaneous capture and measurement for ten different EV biomarkers. When compared for a small test set of human patient plasma, differences observed in levels of a subset of EV biomarkers enabled discrimination between AD+, MCI, and HC subjects. To our knowledge, this is the first study to show combined isolation and analysis of plasma EVs that enables simultaneous comparison of multiple brain-related EV biomarkers for detecting the presence of AD.

Recent studies have shown that brain biomarkers derived from plasma EVs provide much more sensitive detection than those derived from whole plasma. For example, differences in patient levels of the neuropathological marker TDP-43 can be detected in plasma EVs, but not in EV-depleted plasma or in CSF.^25,30 Both AD and Parkinson's disease may be detected by evaluating nEV-bound α-synuclein, Aβ₄₂, total Tau, and pTau181.²⁰ However, because purifying EVs from biofluids is complicated by their small size and high buoyant density, these studies have relied on multi-step, cumbersome hands-on processes for initial isolation of the EVs from blood samples.

We developed a method that utilizes the biophysical properties of EVs to collect them onto a microelectrode array chip directly from biological fluids, such as plasma. Application of AC electrokinetic forces to the suspended plasma particles causes rapid oscillations in polarity, such that particles within a size range of about 30 nm-500 nm experience a net positive movement towards the electrodes. While many similar devices have been developed for isolation of EVs, most of these have not been shown to work for biological fluids, are extremely low-throughput, and/or rely on affinity isolation.^15,31 The ACE capture method shown here does not depend on affinity or type of particle, resulting in unbiased capture of all EVs within the particular size range. Notably, no added pre-processing steps, e.g. filtering, for plasma or serum samples are needed prior to chip capture and analysis.

The biomarkers examined here span the continuum of AD. pTau231, for example, has been reported to identify AD at earlier clinical stages than does pTau181, before deposition of Aβ is sufficient for Aβ PET positivity.^4,12,32 Phosphorylation of Tau at serine 396 can trigger conformational changes, leading to the aggregation that is a hallmark of later-stage AD.^33,34 Increased Aβ₄₀ and Aβ₄₂ levels signify the presence of amyloid plaques in the brain that are the basis for clinical diagnosis of the disease. Other markers, such as α-synuclein and TDP-43, form abnormal aggregates in AD, but may also reflect the presence of other neuropathologies including Parkinson's disease, amyotrophic lateral sclerosis, or LATE.^25,32 It is important to note that while both NCAM and L1CAM have been employed to effect affinity isolation of neuronal EVs, expression of both proteins has been shown to occur also outside the central nervous system.^21,35 This suggests that their utility in isolation of purely neuronal EVs may be limited, and further emphasizes the need for new methods, such as ACE used here, that employ nonselective capture of plasma EVs, with specificity added in the biomarker detection step.

Our results show clear differentiation of pooled AD patient plasma from healthy donor plasma for each of the ten biomarkers examined. As a proof-of-concept, a small set of clinical AD, MCI, patient or HC samples was tested in the OmniVerita system using NCAM1, α-synuclein, pTau231, and TDP-43. As expected, NCAM1 was present in all cohorts, likely reflecting the presence of neuronal or related EVs in all samples. EV-based pTau231 performed well in the ACE microchip system for discriminating AD or MCI from HC, as might be expected based on previous results for pTau231 using whole plasma (p = 0.0300).¹² α-Synuclein could also discriminate AD from the other groups, with EV levels higher in the HC and MCI cohorts (p = 0.0148). While TDP-43 was not statistically different among the three small cohorts, the assay did reveal heterogeneity among individuals that underlines the need for precision medicine in AD. This suggests that future, larger studies should investigate additional qualifiers for sample inclusion. It is interesting to note that the non-AD subjects who scored high in pTau231, TDP-43, or both had normal Aβ_42/40 Lumipulse ratios, but the individual values were elevated (highest pTau231 control sample, Aβ₄₂= 1135, Aβ₄₀= 11,595; highest TDP-43 control sample, Aβ₄₂= 1291, Aβ₄₀= 13,971). One control sample with both high pTau231 and TDP43 was Aβ_42/40 Lumipulse positive. The long-term follow-up for these donors is unknown, but the ability of the ACE AD assay to detect signs of AD in cognitively normal individuals suggests potential utility in early detection of AD.

When analyzed as a multivariate assay, results show the potential for developing assays based on combinations of markers and highlights the capability of the system to measure multiple EV biomarkers in tandem. Future tests could utilize machine learning to analyze multiple biomarker measurements, thus generating an algorithm to detect AD at early stages. Overall, these results support the premise that circulating plasma EVs contain concentrations of multiple brain-derived, AD-related biomarkers, and that effective and unbiased isolation of EVs away from the other contaminating moieties that constitute plasma allows for much better detection of these markers.

When compared to existing commercial ELISA kits designed to measure soluble plasma, serum, or CSF markers, the ACE microchip system for measuring circulating EVs has superior sensitivity. NCAM1 EV levels are in range of the NCAM1 ELISA kit, but values measured by OmniVerita for the other three markers, TDP-43, α-synuclein, and pTau231, were all below the range of reliable detection of the respective ELISA kits (Supplementary Figure 1). This points to the crucial role for EVs in developing sensitive diagnostic assays for AD.

In summary, we show the possibility for developing a simple blood test for early signs of AD or other dementias. The AC electrokinetic microelectrode array chip platform (OmniVerita) facilitates simultaneous capture and quantification of multiple EV biomarkers in situ. The ability to readily monitor changing EV levels of different proteins, such as Tau phosphoforms, suggests potential future utility of the assay in detecting and monitoring AD pathology at pre-clinical stages of the disease or during treatment. Importantly, the use of sensitive blood-based biomarkers for AD diagnosis may increase accessibility to diagnostic services in underserved or rural areas, where specialized laboratories and care are limited.

Supplemental Material

sj-docx-1-alz-10.1177_13872877241291964 - Supplemental material for Single step capture and assessment of multiple plasma extracellular vesicle biomarkers in Alzheimer's disease detection

Supplemental material, sj-docx-1-alz-10.1177_13872877241291964 for Single step capture and assessment of multiple plasma extracellular vesicle biomarkers in Alzheimer's disease detection by Jean M Lewis, Dorathy-Ann Harris, Janell Kosmatka, Emily Mikrut, Jack Evenson, Heath I Balcer, Harmeet Dhani, Juan Pablo Hinestrosa, Robert Rissman and Paul R Billings in Journal of Alzheimer's Disease

Footnotes

Acknowledgments

We appreciate the support and discussions with Dr Jeffrey Dage, Indiana University School of Medicine. We thank ZenBio for the preparation of the UC-purified cell culture EVs. Figures were made with some assistance from Biorender ().

ORCID iDs

Jean M Lewis

Dorathy-Ann Harris

Harmeet Dhani

Author contributions

Jean M Lewis (Conceptualization; Data curation; Formal analysis; Investigation; Methodology; Supervision; Writing – original draft; Writing – review & editing); Dorathy-Ann Harris (Data curation; Formal analysis; Investigation; Methodology; Supervision); Janell Kosmatka (Data curation; Formal analysis; Investigation; Supervision); Emily Mikrut (Data curation; Investigation; Methodology); Jack Evenson (Data curation; Investigation); Heath I Balcer (Conceptualization; Funding acquisition; Project administration; Writing – review & editing); Harmeet Dhani (Project administration; Writing – review & editing); Juan Pablo Hinestrosa (Funding acquisition; Writing – review & editing); Robert Rissman (Methodology; Resources; Writing – review & editing); Paul R Billings (Funding acquisition; Project administration; Resources; Writing – review & editing).

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: All work was performed with the support of the Alzheimer's Drug Discovery Foundation, Grant #GDAPB-201909-2019520, and NIH AG079303 to RR.

Declaration of conflicting interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: JML, DH, JK, EM, JE, HB, HD, JPH, and PB are former employees of Biological Dynamics, which produced the microarray chips used in this study. RR has research support from the National Institute on Ageing, the Alzheimer's Association and is a consultant for Amydis Inc, Bioivt, Lexeo, Keystone Bio, Allyx, DiamiR, Ionis, NeuroQuest and PrecisionMed.

Data availability

The data that support the findings of this study are available from the corresponding author, upon reasonable request.

Supplemental material

Supplemental material for this article is available online.

References

Hampel

Mattke

, et al. Designing the next-generation clinical care pathway for Alzheimer's disease. Nat Aging 2022; 2: 692–703.

Hansson

Blennow

Zetterberg

, et al. Blood biomarkers for Alzheimer's disease in clinical practice and trials. Nat Aging 2023; 3: 506–519.

Janelidze

Bali

Ashton

, et al. Head-to-head comparison of 10 plasma phospho-tau assays in prodromal Alzheimer's disease. Brain 2023; 146: 1592–1601.

Gonzalez-Ortiz

Kac

Brum

, et al. Plasma phospho-tau in Alzheimer's disease: towards diagnostic and therapeutic trial applications. Mol Neurodegener 2023; 18: 18.

Gunes

Aizawa

Sugashi

, et al. Biomarkers for Alzheimer's disease in the current state: a narrative review. Int J Mol Sci 2022; 23: 4962.

Gonzalez-Ortiz

Turton

Kac

, et al. Brain-derived tau: a novel blood-based biomarker for Alzheimer's disease-type neurodegeneration. Brain 2023; 146: 1152–1165.

Hardy-Sosa

Leon-Arcia

Llibre-Guerra

, et al. Diagnostic accuracy of blood-based biomarker panels: a systematic review. Front Aging Neurosci 2022; 14: 683689.

Klyucherev

Olszewski

Shalimova

, et al. Advances in the development of new biomarkers for Alzheimer's disease. Transl Neurodegener 2022; 11: 25.

Mila-Aloma

Ashton

Shekari

, et al. Plasma p-tau231 and p-tau217 as state markers of amyloid-beta pathology in preclinical Alzheimer's disease. Nat Med 2022; 28: 1797–1801.

10.

Palmqvist

Janelidze

Quiroz

, et al. Discriminative accuracy of plasma phospho-tau217 for Alzheimer disease vs other neurodegenerative disorders. JAMA 2020; 324: 772–781.

11.

Planche

Bouteloup

Pellegrin

, et al. Validity and performance of blood biomarkers for Alzheimer disease to predict dementia risk in a large clinic-based cohort. Neurology 2023; 100: e473–e484.

12.

Ashton

Pascoal

Karikari

, et al. Plasma p-tau231: a new biomarker for incipient Alzheimer's disease pathology. Acta Neuropathol 2021; 141: 709–724.

13.

Smirnov

Ashton

Blennow

, et al. Plasma biomarkers for Alzheimer's disease in relation to neuropathology and cognitive change. Acta Neuropathol 2022; 143: 487–503.

14.

Rissman

Langford

Raman

, et al. Plasma Abeta42/Abeta40 and phospho-tau217 concentration ratios increase the accuracy of amyloid PET classification in preclinical Alzheimer's disease. Alzheimers Dement 2024; 20: 1214–1224.

15.

Jin

, et al. Nanosensor-driven detection of neuron-derived exosomal Abeta(42) with graphene electrolyte-gated transistor for Alzheimer's disease diagnosis. Anal Chem 2023; 95: 5719–5728.

16.

Tian

Stewart

Hong

, et al. Blood extracellular vesicles carrying synaptic function- and brain-related proteins as potential biomarkers for Alzheimer's disease. Alzheimers Dement 2023; 19: 909–923.

17.

Couch

Buzas

Di Vizio

, et al. A brief history of nearly EV-erything - the rise and rise of extracellular vesicles. J Extracell Vesicles 2021; 10: e12144.

18.

Mustapic

Eitan

Werner

Jr , et al. Plasma extracellular vesicles enriched for neuronal origin: a potential window into brain pathologic processes. Front Neurosci 2017; 11: 278.

19.

Zhao

Sheng

Wang

, et al. Astrocyte-derived extracellular vesicles: a double-edged sword in central nervous system disorders. Neurosci Biobehav Rev 2021; 125: 148–159.

20.

Blommer

Pitcher

Mustapic

, et al. Extracellular vesicle biomarkers for cognitive impairment in Parkinson's disease. Brain 2023; 146: 195–208.

21.

Gomes

Witwer

. L1CAM-associated extracellular vesicles: a systematic review of nomenclature, sources, separation, and characterization. J Extracell Biol 2022; 1: e35.

22.

Cai

Pang

Wang

, et al. Proteomic profiling of circulating plasma exosomes reveals novel biomarkers of Alzheimer's disease. Alzheimers Res Ther 2022; 14: 181.

23.

Eren

Leoutsakos

Troncoso

, et al. Neuronal-derived EV biomarkers track cognitive decline in Alzheimer's disease. Cells 2022; 11: 436.

24.

Song

Deng

, et al. Brain derived exosomes are a double-edged sword in Alzheimer's disease. Front Mol Neurosci 2020; 13: 79.

25.

Winston

Sukreet

Lynch

, et al. Evaluation of blood-based, extracellular vesicles as biomarkers for aging-related TDP-43 pathology. Alzheimers Dement (Amst) 2022; 14: e12365.

26.

Hinestrosa

Kurzrock

Lewis

, et al. Early-stage multi-cancer detection using an extracellular vesicle protein-based blood test. Commun Med (Lond) 2022; 2: 29.

27.

Hinestrosa

Searson

Lewis

, et al. Simultaneous isolation of circulating nucleic acids and EV-associated protein biomarkers from unprocessed plasma using an AC electrokinetics-based platform. Front Bioeng Biotechnol 2020; 8: 581157.

28.

Lewis

Vyas

Qiu

, et al. Integrated analysis of exosomal protein biomarkers on alternating current electrokinetic chips enables rapid detection of pancreatic cancer in patient blood. ACS Nano 2018; 12: 3311–3320.

29.

Lasser

Eldh

Lotvall

. Isolation and characterization of RNA-containing exosomes. J Vis Exp 2012; 59: e3037.

30.

Zhang

You

, et al. Comprehensive characterization of human brain-derived extracellular vesicles using multiple isolation methods: implications for diagnostic and therapeutic applications. J Extracell Vesicles 2023; 12: e12358.

31.

Sharma

Sheth

Poling

, et al. Rapid purification and multiparametric characterization of circulating small extracellular vesicles utilizing a label-free lab-on-a-chip device. Sci Rep 2023; 13: 18293.

32.

Coomans

Verberk

IMW

Ossenkoppele

, et al. A head-to-head comparison between plasma pTau181 and Tau PET along the Alzheimer's disease continuum. J Nucl Med 2023; 64: 437–443.

33.

Mazzarino

Perez-Corredor

Vanderleest

, et al. APOE3 Christchurch modulates tau phosphorylation and beta-catenin/Wnt/Cadherin signaling in induced pluripotent stem cell-derived cerebral organoids from Alzheimer's cases. bioRxiv 2023. doi: https://doi.org/10.1101/2023.01.11.523290 [Preprint]. Posted January 13, 2023.

34.

Overk

Fiorini

Babolin

, et al. Modeling Alzheimer's disease related phenotypes in the Ts65Dn mouse: impact of age on Abeta, Tau, pTau, NfL, and behavior. Front Neurosci 2023; 17: 1202208.

35.

Dutta

Hornung

Taha

, et al. Biomarkers for parkinsonian disorders in CNS-originating EVs: promise and challenges. Acta Neuropathol 2023; 145: 515–540.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.16 MB