WO2025122650A1

WO2025122650A1 - Extracellular vesicle-based biomarker detection

Info

Publication number: WO2025122650A1
Application number: PCT/US2024/058511
Authority: WO
Inventors: Nicholas Ho; Rasheed SAMAT; Gladys Ho; Rei Nakamoto; Huilin Shao; Peter Maimonis
Original assignee: Sunbird Bio Inc
Current assignee: Sunbird Bio Inc
Priority date: 2023-12-04
Filing date: 2024-12-04
Publication date: 2025-06-12
Anticipated expiration: 2026-06-04

Abstract

This application provides methods and compositions related to detecting extracellular vesicles (EVs) and EV-associated forms of biomarkers for diagnosing a subject. Specifically, this application provides methods for producing disease classifiers from EV-associated biomarkers for diagnosing a subject and subsequently treating a subject diagnosed with a disease.

Description

EXTRACELLULAR VESICLE-BASED BIOMARKER DETECTION

BACKGROUND OF THE INVENTION

Extracellular vesicles (EVs) are nanoscopic, heterogeneous, lipid-rich particles that carry a multitude of cargo biomolecules, including proteins, nucleic acids, and metabolites. These vesicles play essential roles in intercellular communication, facilitating the transfer of bioactive molecules between cells. EVs are involved in various physiological processes, such as immune response modulation, tissue repair, and maintenance of cellular homeostasis. Additionally, they have been implicated in the pathogenesis of multiple diseases, including cancer, neurodegenerative disorders, and cardiovascular diseases.

Historically, EVs were regarded as cellular debris with no intrinsic value. As of late, a growing interest in the fundamental understanding of EV biogenesis has led to the realization that EVs facilitate intercellular communication and are sources of biomarkers in liquid biopsies. However, analytical methods used for diagnosing a patient using EVs have been challenging due to isolation methods generally resulting in contamination from other molecules within the sample. The lack of standardization, data from large patient cohorts, and methodological challenges still hamper current techniques for detecting, assessing, or monitoring diseases.

SUMMARY OF THE INVENTION

The methods and compositions of the disclosure provide detection of extracellular vesicles (EVs), EV markers, and EV-associated forms of biomarkers for diagnosing a subject. Specifically, this application provides methods for producing disease classifiers from EV-associated biomarkers for diagnosing a subject and subsequently treating a subject diagnosed with a disease.

In a first aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including a) providing a sample including a mixture of EVs obtained from the subject, where the mixture of EVs includes an EV population having a surface cargo including an EV-associated form of a disease biomarker; b) using a first binding agent that preferentially binds to the EV-associated form of a disease biomarker cargo to measure the presence of, or level of, the EV population in the sample; c) measuring a total level of all EVs present in the mixture of EVs; and d) on the basis of steps (b) and (c), diagnosing the subject.

In some embodiments, the total level of all EVs present in the mixture of EVs is measured using light scattering, atomic force microscopy, scanning electron microscopy, flow cytometry, surface plasmon resonance, biolayer interferometry, immunoassays, and/or lipid/protein staining.

In some embodiments, i) the EV-associated form of the disease biomarker includes Ap, Ap 42, GFAP, and/or Tau; and ii) the disease is an amyloidosis. In some embodiments, i) the EV-associated form of the disease biomarker includes GFAP, NfL, NSE, and/or S100B; and ii) the disease is a brain vascular damage. In some embodiments, i) the EV-associated form of the disease biomarker includes aSyn, paSyn129, GFAP, and/or NfL; and ii) the disease is a synucleinopathy. In some embodiments, i) the EV- associated form of the disease biomarker includes Tau, pTau231 , and/or pTau396; and ii) the disease is a tauopathy. In some embodiments, i) the EV-associated form of the disease biomarker includes aSyn, paSyn, GFAP, NfL, TDP43, and/or pTDP43; and ii) the disease is a TDP43 proteinopathy. In a second aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including a) providing a sample including a mixture of EVs obtained from the subject, where the mixture of EVs includes a first EV population having a first surface cargo including an EV- associated form of a first marker cargo and a second EV population having a second surface cargo including an EV-associated form of a second marker cargo; b) using a first binding agent that preferentially binds to the EV-associated form of the first marker cargo to measure the level of the first EV population in the sample; c) using a second binding agent that preferentially binds to the EV-associated form of the second marker cargo to measure the level of the second EV population in the sample; and d) on the basis of steps (b) and (c), diagnosing the subject, where the first EV population and the second EV population are different.

In some embodiments, the second binding agent is a first pan binding agent and the level of the second EV population in the sample is a total level of all EVs present in the mixture of EVs, where the first pan binding agent specifically binds a pan marker.

In some embodiments, step c) further includes using a third binding agent that preferentially binds to an EV-associated form of a third marker cargo to measure the level of the second EV population in the sample, where the third binding agent is a second pan binding agent, where the second pan binding agent specifically binds a pan marker. In some embodiments, the first pan binding agent and the second pan binding agent are different. In some embodiments, the pan marker is CD9 or CD81 . In some embodiments, step d) further includes calculating the value of step b) normalized to the value of step c). In some embodiments, step d) further includes calculating the value of step c) normalized to the value of step b). In some embodiments, i) the EV-associated form of the first marker cargo includes Ap, Ap 42, GFAP, and/or Tau; and ii) the disease is an amyloidosis. In some embodiments, i) the EV-associated form of the first marker cargo includes GFAP, NfL, NSE, and/or S100B; and ii) the disease is a brain vascular damage. In some embodiments, i) the EV-associated form of the first marker cargo includes aSyn, paSyn129, GFAP, and/or NfL; and ii) the disease is a synucleinopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes Tau, pTau231 , and/or pTau396; and ii) the disease is a tauopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes TDP43 and/or pTDP43; and ii) the disease is a TDP43 proteinopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and ii) the disease is a brain vascular damage. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and ii) the disease is a synucleinopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and ii) the disease is a tauopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or SYP; and ii) the disease is a TDP43 proteinopathy.

In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; ii) the EV-associated form of the second marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and iii) the disease is a brain vascular damage. In some embodiments, i) the EV- associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; ii) the EV- associated form of the second marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and iii) the disease is a synucleinopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; ii) the EV-associated form of the second marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and iii) the disease is a tauopathy. In some embodiments, i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; ii) the EV- associated form of the second marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and iii) the disease is a TDP43 proteinopathy. In some embodiments, step c) further includes using a third binding agent that preferentially binds to an EV-associated form of a third marker cargo to measure the level of the second EV population in the sample, where the third binding agent is a second pan binding agent, where the second binding agent specifically binds a pan marker.

In some embodiments, step b) further includes using a fourth binding agent that preferentially binds to an EV-associated form of a fourth marker cargo to measure the level of the first EV population in the sample, where the fourth binding agent is a third pan binding agent, where the fourth binding agent specifically binds a pan marker. In some embodiments, the pan marker is CD9 or CD81 .

In a third aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, where the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) using a first binding agent that preferentially binds to the source marker cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) using a second binding agent that preferentially binds to a disease biomarker cargo on the EV subpopulation complexed according to the source marker cargo; d) measuring the presence of, or level of, the disease biomarker cargo on the EV subpopulation complexed according to the source marker cargo; and e) on the basis of step d), diagnosing the subject.

In some embodiments, step c) further includes permeabilizing the EV subpopulation complexed according to source marker with a mild nonionic surfactant to expose disease biomarker cargo on and in the EV subpopulation; and step d) further includes measuring the level of disease biomarker cargo on and in the EV subpopulation complexed according to source marker. In some embodiments, the surfactant is a polysorbate surfactant (e.g., Tween 20®, Tween 40®, Tween 60®, Tween 80®), a polyethylene glycol alkyl ether (e.g., BRIJ® 020), an alkylphenol ethoxylate (e.g., Triton™ X-100, Triton™ X-114, and/or IGEPAL®), or any other nonionic surfactant described herein.

In some embodiments, x) the disease marker cargo includes NSE and/or SWOB; y) the source marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a brain vascular damage. In some embodiments, x) the disease marker cargo includes aSyn, paSyn129, and/or GFAP; y) the source marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a synucleinopathy. In some embodiments, x) the disease marker cargo includes Tau, pTau231 , and/or pTau396; y) the source marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a tauopathy. In some embodiments, x) the disease marker cargo includes TDP43 and/or pTDP43; y) the source marker cargo includes SYP, GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a TDP43 proteinopathy.

In some embodiments, the method further includes (f) using the first binding agent that preferentially binds to the source marker cargo on the EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker; (g) using a third binding agent that preferentially binds to a pan biomarker cargo on the EV subpopulation complexed according to source marker; (h) measuring the level of pan marker on the EV subpopulation complexed according to the source marker; (i) on the basis of steps (d) and (h), diagnosing the subject. In some embodiments, step (i) further includes calculating the value of step (d) normalized to the value of step (h). In some embodiments, step (i) further includes calculating the value of step (h) normalized to the value of step (d). In some embodiments, the pan marker is CD81 or CD9.

In some embodiments, the method further includes (x) measuring the level of the disease biomarker NSE and/or SWOB on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56, (y) measuring the level of the disease biomarker NSE and/or S100B on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56, (z) on the basis of steps (x) and (y), diagnosing the subject with a brain vascular damage, where the first EV subpopulation and the second EV subpopulation are different. In some embodiments, step (z) further includes calculating the value of step (x) normalized to the value of step (y).

In some embodiments, the method further includes (x) measuring the level of the disease biomarker aSyn, paSyn129, and/or GFAP on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56; (y) measuring the level of the disease biomarker aSyn, paSynl 29, and/or GFAP on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56; (z) on the basis of steps (x) and (y), diagnosing the subject with a synucleinopathy, where the first EV subpopulation and the second EV subpopulation are different. In some embodiments, step (z) further includes calculating the value of step (x) normalized to the value of step (y).

In some embodiments, the method further includes (x) measuring the level of the disease biomarker Tau, pTau231 , and/or pTau396 on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56; (y) measuring the level of the disease biomarker Tau, pTau231 , and/or pTau396 on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56; (z) on the basis of steps (x) and (y), diagnosing the subject with a tauopathy, where the first EV subpopulation and the second EV subpopulation are different. In some embodiments, step (z) further includes calculating the value of step (x) normalized to the value of step (y).

In some embodiments, the method further includes (x) measuring the level of the disease biomarker TDP43 and/or pTDP43 on the first EV subpopulation complexed according to the source marker SYP, GLAST, NrCAM, CD171 , and/or CD56; (y) measuring the level of the disease biomarker TDP43 and/or pTDP43 on a second EV subpopulation complexed according to the source marker SYP, GLAST, NrCAM, CD171 , and/or CD56; (z) on the basis of steps (x) and (y), diagnosing the subject with a TDP43 proteinopathy, where the first EV subpopulation and the second EV subpopulation are different. In some embodiments, step (z) further includes calculating the value of step (x) normalized to the value of step (y).

In a fourth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject; b) measuring a level of a first EV population in the mixture of EVs; c) measuring a level of a second EV population in the mixture of EVs; d) calculating a first composite value of step (b) normalized to the value of step (c); e) measuring a level of a third EV population in the mixture of EVs; f) measuring a level of a fourth EV population in the mixture of EVs; g) calculating a second composite value of step (e) normalized to the value of step (f); h) includes combining each composite value of steps (d) and (g) into an algorithm classifier for use in differentially diagnosing the subject; and i) on the basis of step (h) and the algorithm, diagnosing the subject with the disease, where the first EV population and the second EV population are different, and where the third EV population and the fourth EV population are different.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 4; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 4; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 4; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 4; and v) the disease is an amyloidosis.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 8 or 16; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 8 or 16; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 8 or 16; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 8 or 16; and v) the disease is a brain vascular damage.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 12 or 21 ; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 12 or 21 ; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 12 or 21 ; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 12 or 21 ; and v) the disease is a synucleinopathy.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 26; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 26; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 26; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 26; and v) the disease is a tauopathy.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 31 ; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 31 ; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 31 ; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 31 ; and v) the disease is a TDP43 proteinopathy.

In some embodiments, after step g), the method further includes: g1 ) measuring a level of a fifth EV population in the mixture of EVs; g2) measuring a level of a sixth EV population in the mixture of EVs; and g3) calculating a third composite value of step g1 ) normalized to the value of step g2); and where step h) includes combining each composite value of steps d), g), and g3) into an algorithm classifier for use in differentially diagnosing the subject, where the fifth EV population and the sixth EV population are different.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 5; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 5; iii) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 5; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 5; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 5; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 5; and vii) the disease is an amyloidosis.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 9 or 17; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 9 or 17; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 9 or 17; and iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 9 or 17; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of any one of Tables 9 or 17; and vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of any one of Tables 9 or 17; and vii) the disease is a brain vascular damage.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 27; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 27; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 27; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 27; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 27; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 27; and vii) the disease is a tauopathy.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 32; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 32; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 32; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 32; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 32; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 32; and vii) the disease is a TDP43 proteinopathy.

In some embodiments, after step g3), the method further includes: g4) measuring a level of a seventh EV population in the mixture of EVs; g5) measuring a level of an eighth EV population in the mixture of EVs; and g6) calculating a third composite value of step g4) normalized to the value of step g5); and where step h) includes combining each composite value of steps d), g), g3), and g6) into an algorithm classifier for use in differentially diagnosing the subject, where the seventh EV population and the eighth EV population are different.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 18; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 18; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 18; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 18; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 18; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 18; vii) the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 18; viii) the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 18; and ix) the disease is a brain vascular damage.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 23; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 23; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 23; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 23; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 23; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 23; vii) the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 23; viii) the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 23; and ix) the disease is a synucleinopathy.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 28; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 28; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 28; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 28; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 28; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 28; vii) the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 28; viii) the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 28; and ix) the disease is a tauopathy.

In some embodiments, i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 33; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 33; Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 33; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 33; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 33; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 33; vii) the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 33; viii) the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 33; and ix) the disease is a TDP43 proteinopathy.

In some embodiments, the algorithm includes a non-linear model, where the non-linear model includes a multiple logistic regression classifier, a support vector machine, and/or a random forest. In some embodiments, the non-linear model further includes feature selection algorithms, where the feature selection algorithms include a forward selection, a recursive feature elimination, and/or a penalized regression algorithm.

In embodiments of any preceding aspect, the sample is substantially free of exogenous EVs. In embodiments of any preceding aspect, prior to step (b), the method further includes diluting the sample with a diluent, where the diluent includes a protein. In some embodiments, the protein includes IgG, IgA, and/or IgM to competitively inhibit non-specific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent. In some embodiments, the diluent further includes a polymer to increase the preferential binding in steps (b) and (c) by altering the viscosity of the sample and/or inducing the macromolecular crowding effect in the sample, where the polymer includes polyethylene glycol, polyvinylpyrrolidone, dextran, mannitol, betaine, mannitol, sorbitol, xylitol, or other commonly known and used stabilizers. In some embodiments, the diluent further includes a preservative to maintain a long-term sterility of the sample, where the preservative includes any of sodium azide, ProClin™, thimerosal, sodium benzoate, or other commonly known and used preservatives. In some embodiments, the diluent further includes a detergent (i.e. , surfactant) to substantially reduce non-specific binding to a surface.

In embodiments of any preceding aspect, the method further includes diluting the sample with a diluent, where the diluent is substantially free of exogenous EVs. In some embodiments, prior to measurements, no exosomal extraction is performed (e.g., lysing, e.g., lysing by way of detergent or repeated freeze-thaw cycles of the sample). All samples include intact EVs, such that EV-associated biomarkers and EV markers are complexed with the EV.

In a fifth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) complexing a first binding agent that preferentially binds to the source marker cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) permeabilizing the EV subpopulation complexed according to the source marker with a mild nonionic surfactant to expose the disease biomarker cargo on the surface of and/or within the EV subpopulation complexed according to the source marker cargo; d) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the source marker cargo; e) measuring the presence of, or level of, the disease biomarker cargo in the permeabilized EV subpopulation complexed according to the source marker cargo; and f) on the basis of step e), diagnosing the subject.

In a sixth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a first cargo, and (ii) an EV-associated form of a second cargo; b) complexing a first binding agent that preferentially binds to the first cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the first cargo; c) permeabilizing the EV subpopulation complexed according to the first cargo with a mild nonionic surfactant to expose the second cargo on the surface of and/or within the EV subpopulation complexed according to the first cargo; d) complexing a second binding agent that preferentially binds to the second cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the first cargo; e) measuring the presence of, or level of, the second cargo in the permeabilized EV subpopulation complexed according to the first cargo; and f) on the basis of step e), diagnosing the subject.

In a seventh aspect, the disclosure features a method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface including the EV-associated biomarker cargo; b) diluting the sample with a diluent including a protein, wherein the diluent is substantially free of exogenous EVs; c) following step (b), capturing the EV subpopulation having the surface including the EV-associated biomarker cargo; d) washing the surface with a mild nonionic surfactant to permeabilize the captured EV subpopulation; e) using a binding agent that preferentially binds to the EV- associated biomarker cargo on the surface of and/or within the captured and permeabilized EV subpopulation; and f) measuring the presence of, or level of, the EV-associated biomarker cargo.

In an eighth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) diluting the sample with a diluent including a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further includes a mild nonionic surfactant to permeabilize the first EV subpopulation; c) complexing a first binding agent that preferentially binds to the source marker cargo on the surface of and/or within the permeabilized first EV subpopulation from the mixture of EVs in the sample to form a permeabilized EV subpopulation complexed according to the source marker cargo; d) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the source marker cargo; e) measuring the presence of, or level of, the disease biomarker cargo in the permeabilized EV subpopulation complexed according to the source marker cargo; and f) on the basis of step e), diagnosing the subject.

In a ninth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a first cargo, and (ii) an EV-associated form of a second cargo; b) diluting the sample with a diluent including a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further includes a mild nonionic surfactant to permeabilize the first EV subpopulation; c) complexing a first binding agent that preferentially binds to the first cargo on the surface of and/or within the permeabilized first EV subpopulation from the mixture of EVs in the sample to form a permeabilized EV subpopulation complexed according to the first cargo; d) complexing a second binding agent that preferentially binds to the second cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the first cargo; e) measuring the presence of, or level of, the second cargo in the permeabilized EV subpopulation complexed according to the first cargo; and f) on the basis of step e), diagnosing the subject.

In a tenth aspect, the disclosure features a method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface including the EV- associated biomarker cargo; b) diluting the sample with a diluent including a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further includes a mild nonionic surfactant to permeabilize the EV subpopulation; c) following step (b), capturing the permeabilized EV subpopulation having including the EV-associated biomarker cargo on the surface of and/or within the permeabilized EV subpopulation; d) using a binding agent that preferentially binds to the EV-associated biomarker cargo on the surface of and within the captured and permeabilized EV subpopulation; and e) measuring the presence of, or level of, the EV-associated biomarker cargo.

In any of the six preceding aspects, the mild ionic surfactant is a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate. In some embodiments, the mild ionic surfactant is a polysorbate surfactant selected from polyoxyethylene 20 sorbitan monolaurate, polyoxyethylene (4) sorbitan monolaurate, polyoxyethylene 20 sorbitan monopalmitate, polyoxyethylene 20 sorbitan monostearate; and polyoxyethylene 20 sorbitan monooleate, and wherein the permeabilizing step includes exposing the EV to a solution including from about 0.01 to 0.75% (w/w) polysorbate surfactant. In some embodiments, wherein the mild ionic surfactant is a polyethylene glycol alkyl ether selected from PEG-2 oleyl ether, oleth-2; PEG-3 oleyl ether, oleth-3; PEG-5 oleyl ether, oleth-5; PEG-10 oleyl ether, oleth-10; PEG-20 oleyl ether, oleth-20; PEG-4 lauryl ether, laureth-4; PEG-9 lauryl ether; PEG-23 lauryl ether, laureth-23; PEG- 2 cetyl ether; PEG-10 cetyl ether; PEG-20 cetyl ether; PEG-2 stearyl ether; PEG-10 stearyl ether; Polyoxyethylene (20) oleyl ether; PEG-20 stearyl ether; and PEG-100 stearyl ether, and wherein the permeabilizing step includes exposing the EV to a solution including from about 0.01 to 2.5% (w/w) polyethylene glycol alkyl ether. In some embodiments, the mild ionic surfactant is an alkylphenol ethoxylate selected from polyethylene glycol tert-octylphenyl ether and 2-[4-(2,4,4-trimethylpentan-2- yl)phenoxy]ethanol, and wherein the permeabilizing step includes exposing the EV to a solution including from about 0.1 to 2.5% (w/w) alkylphenol ethoxylate. In some embodiments, the protein includes IgG, IgA, and/or IgM to competitively inhibit non-specific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent.

In an eleventh aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) complexing a first binding agent that preferentially binds to the source marker cargo on the surface of and/or within the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of the EV subpopulation complexed according to the source marker cargo; d) measuring the presence of, or level of, the disease biomarker cargo in the EV subpopulation complexed according to the source marker cargo; and e) on the basis of step d), diagnosing the subject.

In a twelfth aspect, the disclosure features a method of diagnosing disease in a subject in need thereof, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface including (i) an EV-associated form of a first cargo, and (ii) an EV-associated form of a second cargo; b) complexing a first binding agent that preferentially binds to the first cargo on the surface of the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the first cargo; c) complexing a second binding agent that preferentially binds to the second cargo on the surface of the EV subpopulation complexed according to the first cargo; d) measuring the presence of, or level of, the second cargo in the EV subpopulation complexed according to the first cargo; and e) on the basis of step d), diagnosing the subject.

In a thirteenth aspect, the disclosure features a method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method including: a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface including the EV-associated biomarker cargo; b) following step (a), capturing the EV subpopulation having including the EV-associated biomarker cargo on the surface of the EV subpopulation; c) using a binding agent that preferentially binds to the EV-associated biomarker cargo on the surface the captured EV subpopulation; and d) measuring the presence of, or level of, the EV-associated biomarker cargo. In any of the three preceding aspects, the method further includes adding a diluent including a mild nonionic surfactant to permeabilize the EV subpopulation, wherein the step of adding the diluent follows (i) step a); (ii) step b), and/or (iii) step c). In some embodiments, the mild ionic surfactant is a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate. In some embodiments, the mild ionic surfactant is a polysorbate surfactant selected from polyoxyethylene 20 sorbitan monolaurate, polyoxyethylene (4) sorbitan monolaurate, polyoxyethylene 20 sorbitan monopalmitate, polyoxyethylene 20 sorbitan monostearate; and polyoxyethylene 20 sorbitan monooleate, and wherein the permeabilizing step include exposing the EV to a solution including from about 0.01 to 0.75% (w/w) polysorbate surfactant. In some embodiments, the mild ionic surfactant is a polyethylene glycol alkyl ether selected from PEG-2 oleyl ether, oleth-2; PEG-3 oleyl ether, oleth-3; PEG-5 oleyl ether, oleth-5; PEG-10 oleyl ether, oleth-10; PEG-20 oleyl ether, oleth-20; PEG-4 lauryl ether, laureth-4; PEG-9 lauryl ether; PEG-23 lauryl ether, laureth-23; PEG- 2 cetyl ether; PEG-10 cetyl ether; PEG-20 cetyl ether; PEG-2 stearyl ether; PEG-10 stearyl ether; Polyoxyethylene (20) oleyl ether; PEG-20 stearyl ether; and PEG-100 stearyl ether, and wherein the permeabilizing step include exposing the EV to a solution including from about 0.01 to 2.5% (w/w) polyethylene glycol alkyl ether. In some embodiments, the mild ionic surfactant is an alkylphenol ethoxylate selected from polyethylene glycol tert-octylphenyl ether and 2-[4-(2,4,4-trimethylpentan-2- yl)phenoxy]ethanol, and wherein the permeabilizing step includes exposing the EV to a solution including from about 0.1 to 2.5% (w/w) alkylphenol ethoxylate. In some embodiments, wherein the diluent further includes a protein, wherein the protein includes IgG, IgA, and/or IgM to competitively inhibit non-specific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent.

In any of the preceding aspects, the step of measuring the presence of, or level of, the EV population and/or biomarker in the sample includes measuring a detectable signal that is a fluorescent, chemiluminescent, radiological, or colorimetric signal. In some embodiments, the step of measuring involves a direct ELISA, an indirect ELISA, a sandwich ELISA, or a competitive ELISA-based assay.

DEFINITIONS

Unless otherwise defined below, all technical and scientific terms used herein are intended to have the same meaning as commonly understood by one of ordinary skill in the art. References to techniques employed herein are intended to refer to the techniques as commonly understood in the art, including variations on those techniques and/or substitutions of equivalent techniques that would be apparent to one of skill in the art. While the following terms are believed to be well understood by one of ordinary skill in the art, the following definitions are set forth to facilitate an explanation of the presently disclosed subject.

As used herein, the singular forms “a,” “an,” and “the” include plural referents unless the content dictates otherwise. Thus, for example, reference to “an antibody” optionally includes a combination of two or more such molecules and the like.

The use herein of the terms “including,” “comprising,” or “having,” and variations thereof is meant to encompass the elements listed thereafter and equivalents thereof as well as additional elements. Embodiments recited as “including,” “comprising,” or “having” certain elements are also contemplated as “consisting essentially of” and “consisting of those certain elements.” As used herein, “and/or” refers to and encompasses any and all possible combinations of one or more of the associated listed items.

The terms “about” and “approximately” as used herein shall generally mean an acceptable degree of error for the quantity measured, given the nature or precision of the measurements. Exemplary degrees of error are within 20% (%), preferably within 10%, and more preferably, within 5% of a given value or range of values. Any reference to “about X” or “approximately X” specifically indicates at least the values X, 0.95X, 0.96X, 0.97X, 0.98X, 0.99X, 1 .01 X, 1 .02X, 1 .03X, 1 .04X, and 1 .05X. Thus, expressions “about X” or “approximately X” are intended to teach and provide written support for a claim limitation of, for example, “0.98X.” Numerical quantities given herein are approximate unless stated otherwise, meaning that the term “about” or “approximately” can be inferred when not expressly stated. When “about” is applied to the beginning of a numerical range, it applies to both ends.

The term “biomarker,” as used herein, is understood to be an agent or entity whose presence or level correlates with an event of interest. The biomarker may be a cell, a protein, nucleic acid, peptide, glycopeptide, an extracellular vesicle, or combinations thereof. For example, the biomarker is an NSE protein whose presence or level indicates whether a subject suffers from or is at risk of developing cancer.

The term “cargo” refers to biomolecules, including, but not limited to, proteins, nucleic acids, lipids, glycans, and metabolites, that are expressed on the surface of and/or within an EV. EV cargo may serve as an EV marker or EV-associated biomarker (e.g., EV-associated disease biomarker).

The term “level” refers to a measured count or quantity, such as, e.g., a concentration.

The term “substantially free of exogenous EVs” as used herein refers to samples free of exogenous EVs or containing a quantity of exogenous EVs that does not functionally interfere with the diagnostic methods of the invention. For example, a solution containing exogenous EVs below a concentration of approximately 10⁸ exogenous EVs/mL are “substantially free of exogenous EVs.” As such, a sample is substantially free of exogenous EVs when the concentration of exogenous EVs is below 10⁸ exogenous EVs/mL. Exogenous EVs can be avoided through careful sample preparation and the avoidance of diluents containing excessive exogenous EVs. For example, diluent components can be sourced to avoid those utilizing, or derived from, animal sources; or any components derived from animal sources be carefully purified to remove EVs to below the functional interference level.

The term “subject” or “individual” means any animal, including any vertebrate or mammal, particularly a human, and can also be referred to e.g., as an individual or patient.

The term “antibody” includes, but is not limited to, synthetic antibodies, monoclonal antibodies, recombinantly produced antibodies, multispecific antibodies (including bi-specific antibodies), human antibodies, humanized antibodies, chimeric antibodies, single-chain Fvs (scFv), Fab fragments, F(ab') fragments, disulfide-linked Fvs (sdFv) (including bi-specific sdFvs), and anti-idiotypic (anti-ld) antibodies, and epitope-binding fragments of any of the above. The antibodies provided herein may be monospecific, bispecific, trispecific, or of greater multi-specificity. Multispecific antibodies may be specific for different epitopes of a polypeptide or for both a polypeptide and a heterologous epitope, such as a heterologous polypeptide or solid support material.

“Antibody fragments” comprise a portion of an intact antibody, for example, the antigen-binding or variable region of the intact antibody. Examples of antibody fragments include Fab, Fab’, F(ab’)2, and Fv fragments; diabodies; linear antibodies {e.g., Zapata et al., Protein Eng. 8(10): 1057-1062 (1995)); singlechain antibody molecules {e.g., scFv); and multispecific antibodies formed from antibody fragments. Papain digestion of antibodies produces two identical antigen-binding fragments, called “Fab” fragments, each with a single antigen-binding site and a residual “Fc” fragment, a designation reflecting the ability to crystallize readily. Pepsin treatment yields an F(ab’)2 fragment with two antigen-combining sites and is still capable of cross-linking antigens.

The terms “protein” and “polypeptide” are used interchangeably and refer to any polymer of amino acids (dipeptide or greater) linked through peptide bonds or modified peptide bonds. Polypeptides of less than 10-20 amino acid residues are commonly called “peptides.” The polypeptides of the invention may comprise non-peptidic components, such as carbohydrate groups. Carbohydrates and other non-peptidic substituents may be added to a polypeptide by the cell in which the polypeptide is produced and will vary with the type of cell. Polypeptides are defined in terms of their amino acid backbone structures; substituents such as carbohydrate groups are generally not specified but may be present, nonetheless. Amino acid polymers may comprise entirely L-amino acids, entirely D-amino acids, or a mixture of L and D amino acids. The term “protein,” as used herein, refers to either a polypeptide or a dimer {e.g., two) or multimer {e.g., three or more) of single chain polypeptides. The single-chain polypeptides of a protein may be joined by a covalent bond, e.g., a disulfide bond, or non-covalent interactions. The terms “portion” and “fragment” are used interchangeably herein to refer to parts of a polypeptide, nucleic acid, or other molecular construct.

The amino acids in the polypeptides described herein can be any of the 20 naturally occurring amino acids, D-stereoisomers of the naturally occurring amino acids, unnatural amino acids, and chemically modified amino acids. Unnatural amino acids (those that are not naturally found in proteins) are also known in the art, as set forth in, for example, Zhang et al., “Protein engineering with unnatural amino acids,” Curr. Opin. Struct. Biol. 23(4): 581 -87 (2013); Xie et al., “Adding amino acids to the genetic repertoire,” Curr. Opin. Chem. Biol. 9(6): 548-54 (2005); all references cited therein.

As used herein, a chemically modified amino acid refers to an amino acid that has been chemically modified. For example, a side chain of the amino acid can be modified to comprise a signaling moiety, such as a fluorophore or a radiolabel. A side chain can also be modified to form a new functional group, such as a thiol, carboxylic acid, or amino group. Post-translationally modified amino acids are also included in the definition of chemically modified amino acids.

As used herein, the terms “binds specifically to,” “specific for,” “specifically binds to,” “preferentially binds,” and the like, are used interchangeably to refer to the binding to a target {e.g., a biomarker, or a disease process biomarker in particular) is significantly stronger than to a control molecule or the binding is significantly stronger as compared to a non-specific or non-selective interaction. In some embodiments, the terms refer to the binding of a specific form {e.g., an EV-associated form) of the target is significantly stronger than a control form {e.g., a soluble form) of the same target. When used in the context of binding to a specific form of a target, a binding agent that preferentially binds to a specific form of the target must also specifically bind to the target. For example, an antibody that binds specifically to an EV-associated form of the Tau protein will also necessarily bind specifically to the Tau protein as compared to a control protein that is not Tau. Specific binding can be measured, for example, by determining a molecule's binding compared to a control molecule's binding. Specific binding can also be determined by competition with a control molecule similar to the target, such as an excess of non-labeled target. In some embodiments, a binding to the target (or a specific form of the target) is specific when the binding is at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 100%, at least 200%, at least 300%, at least 400%, or at least 500% higher than the corresponding control protein or control form of the target. The term “exclusively binds” indicates that only the target biomarker may bind to the agent. A binding agent that exclusively binds to a target falls in the category of binding agents that preferentially bind to the target compared to a control molecule.

The term “target” refers to a molecule which a binding agent specifically binds to. The target may be a marker of an EV. The target may be a soluble form or an EV-associated of a biomarker (or, e.g., disease process biomarker).

The term “healthy control” refers to a subject known not to be suffering from a disease or a subject not at risk of suffering from a disease. The “control subject” or “control” may also be a healthy subject. The “control subject” may have no signs or symptoms of a disease. The term includes a sample obtained from a control subject. For example, the disease may be cancer, neurological, inflammation, autoimmune, or amyloidosis. The “control subject” may be one having no cancer, inflammation, neurological impairment, autoimmune, or amyloidosis. For example, a healthy control for a neurological assay may include a patient sample wherein the patient has never been diagnosed with or is currently experiencing a neurological condition. For example, for an assay designed for diagnosing or monitoring an autoimmune or inflammation, the healthy control may be a patient sample wherein the patient is not currently or does not have an autoimmune or inflammation disease. For an assay designed for the detection of cancer, a healthy control is a patient sample wherein the patient is not currently experiencing cancer or has been previously diagnosed as cancer-free.

The term “significantly different” or “statistically different” refers to the difference between two measurements being statistically different. For example, the level of the biomarker detected in the test sample and the level of the biomarker in the corresponding healthy control being statistically significant. For example, the difference between the level of the biomarker detected in the test sample and the level of the biomarker in the corresponding healthy control may be at least 20%, at least 30%, at least 40%, at least 50%, at least 60%, at least 70%, at least 80%, at least 100%, at least 200%, at least 300% of the level of the biomarker in the corresponding healthy control.

The term “disease process biomarker” or “disease biomarker” means any marker the level of which in a patient in a diseased state is significantly different from (significantly higher or significantly lower than) the healthy controls.

The term “covalent binding” means a chemical bond that involves the sharing of electrons to form electron pairs between atoms.

The term “non-covalent binding” means a bond between two or more molecules that includes more dispersed variations of electromagnetic interactions between molecules. Non-covalent bonding may also be referred to as intermolecular forces. The term “endogenous extracellular vesicles” or “endogenous EVs” refers to EVs in the test sample itself, for example, a plasma sample from a test subject. Endogenous extracellular vesicles can be obtained from their parent cells upon release as a course of their natural biology. These extracellular vesicles released from the parent cells into the surrounding fluid typically include exosomes, microvesicles, and apoptotic bodies produced via various cellular mechanisms. These vesicles can be collected via methods described herein.

The term “exogenous extracellular vesicles” or “exogenous EVs” refer to EVs from a source that is not the test sample. EVs are isolated from a different sample (e.g., bodily fluid or organs and cells) and added to the test sample. For example, the source can be another subject, animal, or an in vitro cell culture. In some embodiments, the exogenous EVs are synthetically synthesized. In some embodiments, exogenous EVs are obtained from one cell line and introduced into a second cell line that is not the cells from which they are isolated. In some embodiments, exogenous EVs are isolated from a first cell type and later added to a second cell type different from the first one. Synthetically made EVs can include those produced from genetically modified cells, such as cells transfected to express a gene of interest double-layered lipid bilayer vesicles produced via common techniques used by those skilled in the art. For example, synthetic EV production methods include thin-film hydration, organic solvent injection, freeze-thaw extrusion, and dehydration-rehydration methods. See, e.g., Datta, B. et al., Intriguing Biomedical Applications of Synthetic and Natural Cell-Derived Vesicles: A Comparative Overview. ACS Applied Bio. Materials, (2021 ) 4(4) 2863- 2885, herein incorporated by reference in its entirety.

The term “soluble biomarker,” “free biomarker,” “biomarker in free form,” or “biomarker in soluble form” refers to a biomarker that can be detected and measured in biological fluids, such as blood, cerebrospinal fluid, synovial fluid, serum, plasma, or urine, and the biomarker is not bound to the EVs.

The term “EV-associated biomarker” refers to a biomarker that binds directly or indirectly to an EV to form a complex with the EV. The biomarker can directly bind to the EV covalently or non-covalently (Hydrogen bonds, Van Der Waals, and the like) via direct interactions to the lipid membrane or via modifications, such as phosphorylation, glycosylation, of the EVs or integral proteins. The biomarker can also indirectly bind to an EV via interactions with bound nucleic acids or other proteins more directly/indirectly bound to the EV. Examples of EV-associated biomarkers and their interactions with EVs are illustrated in FIG. 1. An EV-associated biomarker also refers to a biomarker that is located within the EV as, e.g., internal cargo.

The term “EV subpopulation” refers to EVs having a specific cellular origin. The EV subpopulation may be identified and isolated by way of an EV marker that is expressed on the surface of the EV. Alternatively, an EV subpopulation may be derived exogenously.

The term “EV marker” refers to a biomarker that is expressed on the surface of an EV. An EV marker may be used to identify an EV and quantify the total number of EVs in a sample. An EV marker may be used to identify and isolate an EV subpopulation within a sample that includes a mixture of EVs.

As used herein, the term “detergent” refers to a surfactant or a mixture of two or more surfactants. In some embodiments, the surfactant is a non-ionic surfactant. The mild nonionic surfactant can be a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate surfactant. As used herein, the terms “effective amount,” “therapeutically effective amount,” and “a “sufficient amount” of a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and/or a neuroprotective agent (e.g., in a subject) described herein refer to a quantity sufficient to, when administered to the subject, including a human, effect beneficial or desired results, including clinical results, and, as such, an “effective amount” or synonym thereto depends on the context in which it is being applied. For example, in the context of treating PD, it is an amount of the agent that reduces the motor and/or cognitive symptoms sufficient to achieve a treatment response as compared to the response obtained without administration of the agent. The amount of a given agent that reduces a symptom of PD, MSA, or DLB will vary depending upon various factors, such as the given agent, the pharmaceutical formulation, the route of administration, the subtype of the pathology (e.g., prodromal PD), the identity of the subject (e.g., age, sex, and/or weight) or host being treated, and the like, but can nevertheless be routinely determined by one of skill in the art. Also, as used herein, a “therapeutically effective amount” of a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and a neuroprotective agent of the present disclosure is an amount which results in a beneficial or desired result in a subject as compared to a control. As defined herein, a therapeutically effective amount a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and a neuroprotective agent of the present disclosure may be readily determined by one of ordinary skill by routine methods known in the art. Dosage regimen may be adjusted to provide the optimum therapeutic response.

A “synucleinopathy” is a disorder characterized by misfolding and/or abnormal accumulation of aggregates of alpha-synuclein in the central nervous system (e.g., in neurons or glial cells). Exemplary, nonlimiting synucleinopathies include PD, dementia with Lewy bodies (DLB), multiple system atrophy (MSA), pure autonomic failure, incidental Lewy body disease, pantothenate kinase-associated neurodegeneration, Alzheimer's disease, Down's Syndrome, Gaucher disease, or the Parkinsonism-dementia complex of Guam.

A “tauopathy” is a disorder characterized by misfolding and/or abnormal accumulation of aggregates of tau in the central nervous system (e.g., in neurons or glial cells). Exemplary, non-limiting tauopathies include Alzheimer’s disease, Progressive supranuclear palsy, Corticobasal degeneration, some forms of Frontotemporal dementia, Chronic traumatic encephalopathy, or Parkinsonism linked to chromosome 17.

A “TDP43 proteinopathy” is a disorder characterized by misfolding and/or abnormal accumulation of aggregates of TDP43 in the central nervous system (e.g., in neurons or glial cells). Exemplary, non-limiting TDP43 proteinopathies include Amyotrophic lateral sclerosis, some forms of Frontotemporal dementia, Limbic-predominant age-related TDP43 encephalopathy, or Perry syndrome.

“Amyloidosis” is a disorder characterized by misfolding and/or abnormal accumulation of aggregates of Amyloid beta in the central nervous system (e.g., in neurons or glial cells). Exemplary, non-limiting amyloidosis include Alzheimer’s Disease or Cerebral amyloid angiopathy.

As used herein, the terms “treat,” “treated,’ or “treating” mean therapeutic treatment wherein the object is to ameliorate symptoms of, or slow down (lessen), an undesired physiological condition, disorder, or disease, or obtain beneficial or desired clinical results. Beneficial or desired clinical results include, but are not limited to, alleviation of symptoms; diminishment of the extent of a condition, disorder, or disease; stabilized (i.e ., not worsening) state of condition, disorder, or disease; delay in onset or slowing of condition, disorder, or disease progression; amelioration of the condition, disorder, or disease state or remission (whether partial or total), whether detectable or undetectable; an amelioration of at least one measurable physical parameter, not necessarily discernible by the patient; or enhancement or improvement of condition, disorder, or disease. Treatment includes eliciting a clinically significant response without excessive levels of side effects. Treatment also includes prolonging survival as compared to expected survival if not receiving treatment.

Other features and advantages of the gastrin analogues described herein will be apparent in the following Detailed Description, the drawings, and the claims.

As used herein, the measured levels are named by way of the two EV markers and/or EV- associated biomarkers, or combination thereof, used to measure the corresponding level. For example, if Marker 1 a and Marker 1 b were utilized for an immunoassay to determine the presence of Marker 1 Ab in the subpopulation of EVs bearing Marker 1 Aa, then the name representing the level is “Markert Aa. Marker 1 Ab” and/or “Markert Aa-Marker 1 Ab”. In Examples 15-18, for instance, “Markerl Aa. Marker 1 Ab” is categorized under “Markers 1 A,” referring to the first measured level. Similarly, “Markerl Ba. Markerl Bb” is categorized under “Markers 1 B,” referring to the second measured level. If a composite value is generated from “Markerl Aa. Markerl Ab” and “Markerl Ba. Markerl Bb,” then the name representing the composite value is “Markerl Aa. Markerl AbxMarkerl Ba. Markerl Bb.” Second composite values utilize “Markers 2A” and “Markers 2B,” while third composite values utilize “Markers 3A” and “Markers 3B” and fourth composite values utilize “Markers 4A” and “Markers 4B”.

As used herein, CD171 .NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, NrCAM.NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD9.NSE refers to a measurement of the presence of NSE in the subpopulation of EVs bearing CD9 (pan EV marker).

As used herein, CD171 .CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, NrCAM. CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing GLAST (astrocyte EV marker). As used herein, CD81 .CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD9.CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing CD9 (pan EV marker).

As used herein, CD171 .S100B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.S100B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, NrCAM.SI 00B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.S100B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .S100B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD9.S100B refers to a measurement of the presence of S100B in the subpopulation of EVs bearing CD9 (pan EV marker).

As used herein, CD171 .aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, NrCAM. aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD171 .paSynl 29 refers to a measurement of the presence of paSynl 29 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56. paSynl 29 refers to a measurement of the presence of paSynl 29 in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, NrCAM. paSynl 29 refers to a measurement of the presence of paSynl 29 in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.paSyn129 refers to a measurement of the presence of paSynl 29 in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .paSynl 29 refers to a measurement of the presence of paSynl 29 in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD171 .GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing CD56 (neural lineage EV marker). As used herein, NrCAM.GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, GLAST.CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .CD9 refers to a measurement of the presence of CD9 in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, NrCAM. Tau refers to a measurement of the presence of Tau in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. pTau396 refers to a measurement of the presence of pTau396 in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. pTau231 refers to a measurement of the presence of pTau231 in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, GLAST.Tau refers to a measurement of the presence of Tau in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.pTau396 refers to a measurement of the presence of pTau396 in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.pTau231 refers to a measurement of the presence of pTau231 in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .Tau refers to a measurement of the presence of Tau in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD81 .pTau396 refers to a measurement of the presence of pTau396 in the subpopulation of EVs bearing CD81 (pan EV marker). As used herein, CD81 .pTau231 refers to a measurement of the presence of pTau231 in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD81 .CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD171 .Tau refers to a measurement of the presence of Tau in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 ,pTau396 refers to a measurement of the presence of pTau396 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 ,pTau231 refers to a measurement of the presence of pTau231 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 .CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD56.Tau refers to a measurement of the presence of Tau in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.pTau396 refers to a measurement of the presence of pTau396 in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.pTau231 refers to a measurement of the presence of pTau231 in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, SYP.paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing SYP (synapse EV marker).

As used herein, SYP.CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing SYP (synapse EV marker). As used herein, NrCAM.paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing SYP (neuronal EV marker).

As used herein, NrCAM.aSyn refers to a measurement of the presence of aSyn in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, NrCAM. NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing NrCAM (neuronal EV marker).

As used herein, CD56. paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.GFAP refers to a measurement of the presence of GFAP in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD56.NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing CD56 (neural lineage EV marker).

As used herein, CD171 .paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 .pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 .TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, CD171 .NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing CD171 (brain EV marker). As used herein, CD171 .CD81 refers to a measurement of the presence of CD81 in the subpopulation of EVs bearing CD171 (brain EV marker).

As used herein, GLAST.paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, GLAST.NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing GLAST (astrocyte EV marker).

As used herein, CD81 .paSyn refers to a measurement of the presence of paSyn in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD81 .pTDP refers to a measurement of the presence of pTDP in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD81 .TDP refers to a measurement of the presence of TDP in the subpopulation of EVs bearing CD81 (pan EV marker).

As used herein, CD81 .NfL refers to a measurement of the presence of NfL in the subpopulation of EVs bearing CD81 (pan EV marker).

BRIEF DESCRIPTION OF THE DRAWINGS

FIGs. 1A-1C are schematic representations of exemplary types of EV biomarker interactions. FIG. 1 A illustrates that EVs act as a sponge for the collection of soluble biomarkers in plasma, blood, or bodily fluids in a human sample to concentrate EVs. The association of the biomarkers to the EVs exposes or configures new conformational epitopes in the biomarkers. FIG. 1 B depicts various endogenous sources of EVs (EVs produced by the test subject's own organs) and exogenous sources of EVs (EVs from subjects other than the test subjects or EVs derived from cell lines in vitro). FIG. 1C depicts exemplary types of EV interactions, including covalent bonding, non-covalent bonding, direct binding (e.g., by modifications (e.g., phosphorylation or glycosylation) or lipid-protein interactions), and indirect binding (e.g., protein-DNA/RNA interactions or protein-protein interactions).

FIG. 2 shows schematic illustrations of various EV-based immunoassay configurations for the detection of a biomarker of interest. Each EV-based immunoassay configuration comprises a pair of affinity agents (i.e., binding agents). The affinity agents used in these configurations generally belong to the one of the four categories as shown in the top panel from left to right: i) affinity agents that specifically bind to the soluble biomarker (“SSA"; regular width antibody), ii) affinity agents that specifically bind to the EV- associated biomarker ("EVbA"; thin width antibody), iii) affinity agents that bind to both soluble and EV- associated biomarkers ("UA"; thick width antibody), and iv) affinity agents that specifically bind to an EV marker ("EVA"; dotted-line antibody). Different combinations of the affinity agents can be designed to detect soluble biomarkers ("Soluble biomarker specific assays"), soluble and EV-associated biomarkers ("soluble and EV-associated biomarkers assays"), EVs ("EV assays"), and EV-associated biomarkers ("EV-associated biomarker specific assays"). Exemplary assay configurations are shown in the left bottom panel for assays designed to detect EV-associated biomarkers within a mixture of EVs and the total EVs within a sample (under "Gen 1 (First Generation) Assay Design"). Other exemplary assays configurations are shown in the right bottom panel for assays designed to detect EV-associated biomarkers within an EV subpopulation and the total EVs within a subpopulation by using at least one EVA (under "Gen 2 (Second Generation) Assay Design").

FIGs. 3A-3C show the how signals vary between soluble, EV-associated, and soluble and EV- associated biomarkers among disease and healthy controls for rare and specific disease relevant biomarkers. FIG. 3A depicts assays for detecting soluble and EV- associated biomarkers for endogenous (white) EVs within a sample and exogenous (black) EVs added to a sample. The top assay configurations in FIG. 3A show that an SSA binds the soluble biomarker whereas EVbAs bind the EV-associated biomarker. The bottom assay configurations in FIG. 3B shows that UAs bind soluble and EV-associated forms of biomarkers. FIG. 3B and FIG. 3C show the results of using the three assays depicted in FIG. 3A to use EVs to enrich for a rare soluble biomarker (e.g., Tau, GFAP, and the like) and to use EV-associated biomarker interaction to recognize specific disease-relevant forms of a common soluble biomarker (e.g., alpha- synuclein, CD171 , and the like), respectively. The pie chart shows the quantity of soluble (dotted) versus EV- associated (white) forms of the biomarkers. The bar graphs of soluble and EV-associated forms of rare biomarkers for diseased (dashed) and healthy (white) patients demonstrate how EV-associated biomarkers provide enhanced detection of rare and disease relevant forms of biomarkers in diseased patients against healthy patients.

FIG. 4 shows the results of detecting known biomarkers and EV markers, CD9, CD63, CD81 , CD56, CD171 , NrCAM, EAAT1 , SYP, Alpha Synuclein, phosphorylated Alpha Synuclein, Tau, phosphorylated Tau, Amyloid beta, Amyloid Beta 42 peptide, TDP43, phosphorylated TDP43, GFAP, NfL, SWOB, NSE, IL-6, IL-1 beta, and TNF-alpha, using soluble biomarker specific assays (vertical gray bars) and the EV-associated biomarker specific assays (vertical white bars). The assays detect the markers in endogenous EVs. Horizontal bars underneath the marker names indicate the specificity of the markers: EV markers (black), neurology-specific biomarkers (diagonally-striped), protein aggregation-specific biomarkers (zig-zag), inflammatory/autoimmune related biomarkers (vertically-striped) and cancer-specific biomarkers (diamond-patterned).

FIGs. 5A and 5B show the signals for EV-associated and soluble forms of abundant and rare biomarkers and for total EVs among 30 patients. FIG. 5A shows the results of using the soluble CD171 specific assay (striped), the EV-associated CD171 specific assay (white), and the EV assay (black) to detect CD171 , which is highly abundant in its soluble form in the blood, in different patient samples. FIG. 5B shows the results of similar assays performed to detect Tau in these patient samples, whose soluble form typically has a low concentration in the blood.

FIG. 6 shows the specificity of the antibody pairs (x-axis) toward recombinant proteins (y-axis) used in brain pathologic, brain function, and EV marker immunoassays, all of which are identified along both axes by a black, gray, or light gray bar, respectively. The bolded black boxes within the grid further highlight the aforementioned groups of recombinant proteins. Recombinant proteins tested included: Amyloid beta-40 (Ap40), Amyloid beta-42 (Ap42), Tau-441 (2N4R), phosphorylated Tau (pTau), GSK3-beta, alpha-synuclein (aSyn), phosphorylated alpha-synuclein (paSyn), PLK2, GFAP, NfL, NSE, S OB, CD9, CD63, CD81 , CD56, CD171 , and NrCAM. The scale bar shows the strength of the signal for measured from the immunoassay with a value of 0.0 being white transitioning to a value of 1 .0 being black. Antibody pairs demonstrated high specificity toward their target recombinant protein with no cross-reactivity toward any other recombinant proteins. Total Ap, Tau, and aSyn assays were able to bind to all isoforms of the protein, while protein form specific assays were specific only to its reported isoform/post-translational modification.

FIGs. 7A and 7B show the specificity of assays with two different configurations toward EV- associated Tau (left) and EV-associated alpha synuclein (right). FIG. 7A-D utilized an EV-associated biomarker ("EVbA"; thin-width antibody) antibody or soluble and EV-associated biomarkers ("UA"; thick-width antibody) antibody as the first binding agent, while an EV marker ("EVA"; dashed-line antibody) was utilized as the second binding agent (y-axis). The opposite configuration was used, with the results shown on the x- axis. The EV markers measured include CD56 and CD81 . The different configurations yielded high correlation, as shown by the trendlines, between the signals obtained from the plasma of 11 subjects for both EV-associated Tau and EV-associated alpha synuclein, demonstrating that the specificity toward the target is not dependent on the configuration.

FIG. 8 shows the sizes of soluble versus EV-associated biomarkers. Human plasma was filtered by passing through size filtration devices (800 nm pore size, 200 nm pore size, or 100 kDa pore size). The levels of endogenous EV marker (CD9), endogenous specially folded EV-associated biomarker, or soluble recombinant protein (Tau or Amyloid beta) spiked into the sample were measured in the retentate (black) (larger than the filter pore size) and filtrate (white) (smaller than filter pore size). The results show that most EV-associated biomarker sizes were larger than 100 kDa (~3 nm) and smaller than 220 nm, which is within the expected size range for most EVs. The results also show that the size of the soluble biomarker was smaller than 100 kDa, also within the expected size range for monomeric Tau protein (55 - 82 kDa).

FIG. 9 are scanning electron microscopy (SEM) micrographs of EV-associated biomarkers Tau and Amyloid beta, captured by EV-associated Tau-specific antibodies and EV-associated Amyloid beta-specific antibodies. The scale bar is equivalent to 1 mm. Particles observed are within the EV size range (50 nm - 300 nm). Soluble biomarker-specific antibody for Tau and Amyloid beta were not detected possibly due to the small size, 2 nm - 5 nm, which are below the resolution of scanning electron microscopy.

FIG. 10 shows a schematic detailing components of the blood matrix that can potentially interfere with direct immunoassays and components of a sample diluent that can cancel out the potentially interfering component effects.

FIG. 11 shows a schematic explaining why direct measurement of isolated EVs and their associated cargo results in much higher signal than attempted lysis and measurement of the lysate. Assays were designed for capture of EV-associated and soluble forms of a biomarker in the supernatant and the precipitate with and without exposure to a lysis agent. In conditions with and without a lysing agent, the obtained biomarker signal was remarkably higher in the precipitate (striped) compared to the supernatant (white), indicating that introduction of a lysing agent does not break down the EV as expected. Without being bound by any theory, it is hypothesized that the EV is permeabilized when exposed to lysing conditions, increasing access to the internal cargo or surface-bound cargo.

FIGs. 12A and 12B show the results of two individuals' EV-associated protein biomarker (amyloid beta; black) and EV marker (CD9; white) levels in human plasma samples. The samples were subjected to multiple freeze-thaw cycles before detection (FIG. 12A). The sample were also incubated with non-ionic surfactant at room temperature for various period of time (FIG. 12B). These results show that EVs and EV- associated biomarkers in blood are stable after various types of common handling and treatment options.

FIG. 13 shows how a short treatment with a non-polar (i.e. , non-ionic) detergent (Tween 20®) greatly increases the signal for two EV cargo proteins, Tau and Amyloid Beta, and their subtypes, pTau and Amyloid Beta 42. The signal is enhanced 100-fold after treatment for at least 5 minutes with Tween 20® compared to the untreated sample (0 minutes), indicating the permeabilization of the EV to increase access to EV- associated biomarkers located on the surface of the EV and internally. Without being bound to any theory, it is hypothesized that a significant portion of the EV-associated biomarkers are located inside the EV, leading to the sharp gain in signal after introduction of the detergent. Additional treatment does increase the signal slightly (up to 20%) but not by a significant amount, showcasing how it is simple to permeabilize the EVs, but difficult to fully lyse it (would expectedly lead to a sharp decrease in signal).

FIGs. 14A and 14B show the use of different detergents as lysing agents for EVs that are immunocaptured on a surface and measurement of EV-associated cargo on the surface post-treatment. FIG. 14A shows the effect of each detergent solution (non-polar (i.e., non-ionic) detergents represented as white bars and polar detergents represented as black bars) on the antibody affinity (both CD81 EV capture and Tau detection antibodies) as determined using a synthetic target. FIG. 14B shows the effect of treating immunocaptured EVs with mild non-polar (i.e., non-ionic) detergent on the Tau cargo signal after correction for the effect of detergent on capture antibody affinity. If EVs would be completely lysed, the signal would be expected to be lower than the untreated EVs, as the tau protein cargo would be released into the lysate/supernatant. Instead, a 100 times stronger signal of tau protein on the surface is achieved postdetergent treatment. This shows that the treatment with a wide range of detergents is able to permeabilize the EV membrane to measure internal cargo, but also that the EVs captured on the surface a resistant to lysis such that most of the cargo is not released into the lysate/supernatant, but instead remains associated with the EV complex on the surface. FIGs. 15A and 15B show the resistance of neural EV-associated alpha synuclein to peripheral tissue contamination. FIG. 15A shows the results of measuring EV-associated alpha synuclein (black) and soluble alpha synuclein (white) after incubation of the plasma samples of two subjects with intact red blood cells (RBCs), semi-lysed RBCs, and lysed RBCs. For both subjects, the signal of soluble alpha-synuclein increased after incubation with semi-lysed RBCs and lysed RBCs, whereas the signals of EV-associated alpha synuclein remained stable. FIG. 15B shows the change in signal of EV-associated alpha synuclein and soluble alpha synuclein from the whole blood of 6 subjects after incubation at room temperature in increments of time. The plasma from each subject passed the hemolysis test. The signals for soluble alpha synuclein increased over time, whereas the signal for EV-associated alpha synuclein remained stable over time. This demonstrates that even trace amounts of RBC lysis can cause significant amounts of alpha synuclein leakage and contaminate the sample, something which the neural EV-associated biomarkers are robust to.

FIG. 16 depicts assay methods for detecting EV-biomarkers dependent on EV subpopulation and biomarker properties. FIG. 16 shows the results of detecting the endogenous circulating biomarkers associated with EVs (EV type 1 ; white) and the exogenous EVs of different types (EV type 2; black) added to the sample for two different biomarkers in a healthy sample as compared to a disease sample. These results demonstrate preferential binding between different types of EV and different patients, indicating the the same biomarkers between two different subjects may have different affinities for the same EV depending on the state of the individual (i.e., diseased vs healthy). These results also demonstrate that other types of assays to measure how the same biomarker state binds differently to different types of EVs (either endogenous, from different cell types, or exogenous from various sources) based on the EV subtype properties can be built utilizing recognition agents specific for different EV subtypes.

FIG. 17 depicts the normalized signal heat map for Tau expression in a cohort of 30 patient samples. Using different EV subpopulation markers (CD9, CD171 , CD56, NrCAM, GLAST), Tau bound to each type of EV in blood at different levels. The value of the normalized signal is expressed by the heat map using a grey scale heat map transitioning from white (value of 0.0) to black (value of 1 .0). In the cohort of 30 patient samples, many different patterns of EV subpopulation bound Tau across different samples, demonstrating the other derived EVs contain different levels of the same biomarker that is biologically different.

FIG. 18 shows the competitive binding between endogenous EVs and exogenous EVs of different types. For a disease control (black bars), the normalized signal of the EV-associated biomarker for the endogenous EV decreased after incubation with the first type of exogenous EV (gray EV) and second type of exogenous EV (black EV). For a healthy control (white bars), the normalized signal of the EV-associated biomarker for the endogenous EV also decreased after incubation with the first type of exogenous EV (grey EV) and second type of exogenous EV (black EV). Between the two subjects, the different exogenous EVs yielded different magnitudes of decrease in the normalized signal of the EV-associated biomarker. Furthermore, for each subject, the different exogenous EVs yielded different magnitudes of decrease in the normalized signal of the EV-associated biomarker. This result demonstrates that introduction of exogenous EVs into a sample with endogenous EVs can cause the dissociation of biomarkers from the endogenous EVs and subsequent association of biomarkers with the exogenous EVs, resulting in different epitopes of the biomarker being exposed on the surface of the exogenous EVs. These changes in the EV-associated biomarker can be used to distinguish between different biomarker properties.

FIG. 19 illustrates using exogenous EVs derived from cell lines added into the same two different samples (A & B) can differentially capture Tau and Amyloid Beta in the blood. This results in different measures by different EV types within the same individual, and/or the same EV type between different individuals. The EV-associated Tau and amyloid beta signals varied between incubating the blood sample with neuronal cell-derived EVs (diagonally-striped), Glial cell-derived EVs (vertically-striped), and Epithelial cell-derived EVs (white), with the responses also being different between the two patients.

FIGs. 20A-20C show how EVs from different cell types (EV subpopulations) influence the binding properties between a biomarker and endogenous EVs using assays designed to be sensitive to an EV- associated Tau epitope in endogenous EVs. Two human plasma samples including endogenous EV- associated Tau were assessed before and after incubation with increasing doses of epithelial cancer EVs (A431 cell line), neuronal EVs (SHSY5Y cell line), cow blood isolated (bovine) EVs, goat blood (caprinae) isolated EVs, and chicken blood (gallus) isolated EVs (FIGs. 20A and 20B). Between the two human plasma samples, the same dose of the same type of EV elicited different responses and increasing doses of the same type of EV elicited different slopes of response (FIG. 20C). In all cases, the EV-associated Tau signal decreased compared to the untreated sample as the sample was incubated with an increasing amount of the different type of EV.

FIGs. 21 A and 21 B illustrate the use of exogenous EVs to concentrate biomarkers in a subject’s blood sample. The amount of EV-associated Tau detected increased as the number of exogenous EVs used in the incubation increased (FIG. 21 A) and as the duration of the incubation increased (FIG. 21 B).

FIGs. 22A and 22B show the results of EV quantity and EV-associated biomarker quantity (Tau, pTau, Amyloid beta, Amyloid beta 42, GFAP, and NfL) measured in blood samples across seven clinical cohorts. The EV quantity within 7 independently collected cohorts ranges over four orders of magnitude (FIG. 22A). The EV quantity positively correlates with the EV-associated biomarker quantity among the samples for all biomarkers measured (FIG. 22B).

FIG. 23 compares the results of FIGs. 7A-D for measures of EV-associated aSyn (square, black points) and EV-associated Tau (circular, black points) using both CD81 -specific antibodies and CD56- specific antibodies as EVA antibodies to the level of total EVs measured by immunoassays with two different EV marker (EVA) antibody pairs: CD81 -specific antibody as the first and second binding agents (top) and CD56-specific antibody replacing CD81 -specific antibody as the first binding agent (bottom). Despite levels of EV-associated biomarkers correlating generally with the levels of total EVs as indicated by the trendlines, there are outliers from the general trend which may reflect differences depending on individual, type of EV cargo, and type of EV (white points).

FIG. 24A and 24B show how the EV-quantity bias can be corrected (e.g., normalized) from EV- associated biomarker quantity by utilizing a regression adjustment. The positive correlation between EV- associated amyloid and EV quantity among amyloid negative (black points) and amyloid positive (white points) controls prevents the classification between disease and control groups (FIG. 24A). As the quantity of EVs in the blood has many causes and is not purely dependent on disease state, the bias from the quantity of EVs can be corrected by utilizing a regression adjustment. The resulting composite values for amyloid negative and positive controls yield a good separation between the two groups (FIG. 24B). These results demonstrate that total EV quantity can be utilized in conjunction with EV-associated biomarker quantities to calculate a composite value to remove the total EV bias from measurement.

FIGs. 25A, 25B, and 25C shows the utility of the correction and combination algorithm in detecting EV associated total amyloid beta and amyloid beta 42 peptide. FIGs. 25A and 25B show plots of the signal for the amyloid positive controls (filled circular points) and healthy controls (filled triangular points), while FIG. 25C shows plots of amyloid aggregation prediction score for the amyloid positive controls (open circular points) and healthy controls (filled circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 25A shows that the raw quantity of EV bound amyloid beta, amyloid beta 42, GFAP, Tau, NfL or total EV quantity does not significantly differ between amyloid PET positive and healthy controls. FIG. 25B shows that the EV quantity adjusted amyloid beta and amyloid beta 42 peptide do have a difference between the amyloid PET positive and healthy controls. FIG. 25C shows that combining the two measures into a multiple logistic regression classifier (with leave-out-out cross validation) produced a classifier for amyloid PET positivity that is better than any individual measure (AUC: 0.87). These methods may extend to classifying other diseases resulting from amyloid beta aggregation.

FIGs. 26A and 26B show the results of the classifier detecting EV-associated amyloid beta and amyloid beta 42 peptide. The peptide levels can be used to classify amyloid buildup status in the brain. FIG. 26A and 26B show a plot on the left of the amyloid aggregation prediction score for healthy (filled circular points) and amyloid positive (open circular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 26A shows that the levels of amyloid beta and amyloid beta 42 peptide bound to EV normalized against total EV count was significantly different in amyloidpositive individuals compared to amyloid-negative individuals in a discovery cohort of 30 individuals (AUC: 0.87). FIG. 26B shows the results from a separate cohort of 15 individuals with similar performance (AUC: 0.84) which verified the results of the discovery cohort. These methods may extend to classifying other diseases resulting from amyloid beta aggregation.

FIGs. 27A, 27B, and 27C show the results of detecting EV-associated Neuron-specific enolase (NSE), EV-associated S100 calcium-binding protein B (S100B), and total EVs for stroke and healthy controls. FIGs. 27A and 27B show plots of the signal for the healthy controls (filled circular points) and stroke controls (open circular points in FIG. 27A and open triangular points in FIG. 27B), while FIG. 27C shows plots of brain vascular damage (BVD) prediction score for the healthy controls (filled circular points) and stroke controls (open circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 27A shows that each individual biomarker is not significantly different between stroke and control populations when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 27B, after regressing out the level of total EVs from the EV-associated biomarker signal there is still little difference between the stroke and healthy controls. However, FIG. 27C shows that using these two measures to build a logistic regression classifier (with leave-out-out cross validation), EV- associated an enhanced brain stroke classifier was obtained (AUC: 0.816). These results show that the level of EV-associated biomarkers can be normalized to the level of total EVs and subsequently combined by way of algorithmic correction to obtain an accurate stroke classifier. These methods may extend to classifying other diseases resulting from brain vascular damage (BVD). FIGs. 28A, 28B, and 28C shows the results of using different EV bound cargo and/or soluble cargo for classifying stroke in a cohort of 22 ischemic stroke and 23 healthy controls (n=45). FIGs. 28A, 28B, and 28C show a plot on the left of the brain vascular damage prediction (BVD) score for healthy (filled circular points) and stroke (open circular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 28A is identical to FIG. 27C, showing how EV bound NSE and S100B can be combined to build an enhanced brain stroke classifier (AUC: 0.816). FIG. 28B shows that the unbound soluble forms of NSE and SWOB measured in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.572). This demonstrates that it is the EV-associated form of these proteins that are relevant to brain vascular damage (BVD). Furthermore, FIG. 28C shows that measures of EV-associated glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.506). This demonstrates that the specific cargo associated with the EV impart important functional information that can be used to reflect disease processes and accurately diagnose disease.

FIGs. 29A, 29B, and 29C shows the results of using different EV bound cargo and/or soluble cargo for classifying stroke in an independently collected cohort of 18 ischemic and hemorrhagic stroke and 32 healthy controls (n=50). FIGs. 29A, 29B, and 29C show a plot on the left of the brain vascular damage prediction (BVD) score for healthy (circular points) and stroke (triangular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 29A shows how EV bound NSE and SWOB can be combined to build an enhanced brain stroke classifier (AUC: 0.803). FIG. 29B shows that the unbound soluble forms of NSE and S100B measured in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.526). This demonstrates that it is the EV- associated form of these proteins that are relevant to brain vascular damage (BVD). Furthermore FIG. 29C shows that measures of EV-associated glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.562). This demonstrates that the specific cargo associated with the EV impart important functional information that can be used to reflect disease processes and accurately diagnose disease.

FIGs. 30A, 30B, and 30C show the results of detecting brain (CD171 ) and astrocyte (GLAST) EV- associated Neuron-specific enolase (NSE), EV-associated S100 calcium-binding protein B (SWOB), and EVs quantities for stroke and healthy controls in a cohort of 12 ischemic stroke and 16 healthy controls (n=28). FIGs. 30A and 30B show plots of the signal for the healthy controls (filled circular points) and stroke controls (open circular points), while FIG. 30C shows plots of brain vascular damage (BVD) prediction score for the healthy controls (filled circular points) and stroke controls (open circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 30A shows that each individual biomarker is not significantly different between stroke and control populations when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 30B, after regressing out the levels of brain or astrocyte EVs from the respective EV-associated biomarker signal there is some difference between the stroke and healthy controls. However, FIG. 30C shows that using these four measures to build a logistic regression classifier (with leave-out-out cross validation), an enhanced brain stroke classifier was obtained (AUC: 0.963). These results show that the level of EV-associated biomarkers can be normalized to the level of total EVs within the corresponding EV subpopulation and subsequently combined by way of algorithmic correction to obtain an accurate stroke classifier. These methods may extend to classifying other diseases resulting from brain vascular damage (BVD).

FIGs. 31A, 31B, and 31C show the results of detecting brain (CD171 ), neuron (NrCAM), astrocyte (GLAST), and total (CD81/CD9) EVs quantities for stroke and healthy controls in a cohort of 12 ischemic stroke and 16 healthy controls (n=28). FIGs. 31 A and 31 B show plots of the signal for the healthy controls (filled circular points) and stroke controls (open circular points), while FIG. 31 C shows plots of brain vascular damage (BVD) prediction score for the healthy controls (filled circular points) and stroke controls (open circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 31 A shows that each individual EV quantity is not significantly different between stroke and control populations when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 31 B, after regressing out the levels of brain or astrocyte EVs from the respective EV-associated biomarker signal there is some difference between the stroke and healthy controls. However, FIG. 31 C shows that using these three measures to build a logistic regression classifier (with leave-out-out cross validation), an enhanced brain stroke classifier was obtained (AUC: 0.988). These results show that the level of different brain EV subpopulation quantities can be normalized to the level of total EVs and subsequently combined by way of algorithmic correction to obtain an accurate stroke classifier. These methods may extend to classifying other diseases resulting from brain vascular damage (BVD).

FIGs. 32A, 32B, and 32C show the difference in performance between using total EV bound NSE and SWOB (FIG. 32A, identical to FIG. 28A) in a cohort of 22 ischemic stroke and 23 healthy controls (n=45), and the brain and astrocyte EV subpopulation bound NSE and SWOB (FIG. 32B, identical to FIG. 30C) or brain, neuron, astrocyte, and total EV quantities (FIG. 32C, identical to FIG. 31C) in a selected subset of the same cohort 12 ischemic stroke, 16 healthy, n=28). FIGs. 32A, 32B, and 32C show a plot on the left of the brain vascular damage prediction (BVD) score for healthy (filled circular points) and stroke (open circular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. The performance of the total EV signature (AUC: 0.816) is not as good as the neural EV subpopulation signatures in FIGs. 32B and 32C (AUC: 0.963 or 0.988, respectively). This demonstrates that neural EV subpopulations in the blood can provide more information more reflective of the brain pathology compared to total EVs in the blood. These methods may extend to classifying other diseases resulting from brain vascular damage.

FIG. 33A, 33B, and 33C again compares the performance of using total EV bound cargo or neural EV subpopulations in an independently collected cohort of 18 ischemic and hemorrhagic stroke and 32 healthy controls (n=50). FIGs. 33A, 33B, and 33C show a plot on the left of the brain vascular damage prediction (BVD) score for healthy (circular points) and stroke (triangular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 33A, identical to FIG. 29A, shows the total EV bound NSE and SWOB has an AUC of 0.803. Once again in this independent cohort, utilizing the brain and astrocyte bound NSE and SWOB (AUC: 0.940) or brain, neuron, astrocyte, and total EV quantities (AUC: 0.924) in blood was able to build a more accurate stroke classifier. This validates that neural EV subpopulations in the blood can provide more information more reflective of the brain pathology compared to total EVs in the blood. These methods may extend to classifying other diseases resulting from brain vascular damage. FIGs. 34A, 34B, and 34C show the results of detecting EV-associated alpha-synuclein, phosphorylated alpha-synuclein, GFAP, NfL, and total EV quantity for Parkinson’s disease controls and healthy controls. FIGs. 34A and 34B show plots of the signal for the healthy controls (filled, black circular points) and Parkinson’s disease controls (open or gray circular points), while FIG. 34C shows plots of alpha synuclein aggregation prediction score for the healthy controls (filled circular points) and Parkinson’s disease controls (open circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 34A shows that each individual biomarker is not significantly different in PD compared to control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 34B, after regressing out the level of total EVs from the EV-associated biomarker signal there is still little difference between the stroke and healthy controls. However, as shown in FIG. 34C, when building a logistic regression classifier (with leave-out-out cross validation) to combine EV-associated alpha synuclein, phosphorylated alpha synuclein, and GFAP, a composite value of these measures created an enhanced PD classifier (AUC: 0.866), better than any individual measure. These results show that the level of EV-associated biomarkers can be normalized to the level of total EVs and subsequently combined by way of algorithmic correction to obtain an accurate Parkinson’s disease classifier. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 35A, 35B, and 35C shows the results of using different EV bound cargo and/or soluble cargo for classifying PD in a cohort of 15 symptomatic Parkinson’s Disease and 25 healthy controls (n=40). FIG. 35A, 35B, and 35C show a plot on the left of the alpha synuclein aggregation prediction score for healthy (filled circular points) and Parkinson’s Disease (open circular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 35A is identical to FIG. 34C, showing how EV bound alpha synuclein, phosphorylated alpha synuclein, and GFAP can be combined to build an enhanced brain stroke classifier (AUC: 0.866). FIG. 35B shows that the unbound soluble forms of alpha synuclein, phosphorylated alpha synuclein, and GFAP measured in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.515). This demonstrates that it is the EV- associated form of these proteins that are relevant to brain vascular damage. Furthermore, FIG. 35C shows that measures of EV-associated glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.539). This demonstrates that the specific cargo associated with the EV impart important functional information that can be used to reflect disease processes and accurately diagnose disease.

FIGs. 36A, 36B, and 36C shows the results of using different EV bound cargo and/or soluble cargo for classifying PD in an independently collected cohort of 18 symptomatic Parkinson’s Disease and 32 healthy controls (n=50). FIGs. 36A, 36B, and 36C show a plot on the left of the alpha synuclein aggregation prediction score for healthy (circular points) and Parkinson’s Disease (triangular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. FIG. 36A shows how EV bound alpha synuclein, phosphorylated alpha synuclein, and GFAP can be combined to build an enhanced brain stroke classifier (AUC: 0.814). FIG. 36B shows that the unbound soluble forms of alpha synuclein, phosphorylated alpha synuclein, and GFAP measured in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.596). This demonstrates that it is the EV- associated form of these proteins that are relevant to brain vascular damage. Furthermore, FIG. 36C shows that measures of EV-associated glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) in the same cohort utilizing the same methods do not accurately classify stroke patients (AUC: 0.518). This demonstrates that the specific cargo associated with the EV imparts important functional information that can be used to reflect disease processes and accurately diagnose disease. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 37A, 37B, and 37C show the results of detecting EV-associated alpha-synuclein, phosphorylated alpha-synuclein, GFAP, and total EV quantity for Parkinson’s disease and healthy controls in a cohort of 14 symptomatic Parkinson’s Disease and 12 healthy controls (n=26). FIGs. 37A and 37B show plots of the signal for the healthy controls (circular points) and Parkinson’s disease controls (diamond points), while FIG. 37C shows plots of alpha synuclein aggregation prediction score for the healthy controls (circular points) and Parkinson’s disease controls (diamond points) and of the ROC curve for determining the performance of the classifier model. FIG. 37A shows that each individual biomarker is not significantly different in PD compared to control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 37B, after regressing out the level of total EVs from the EV-associated biomarker signal there is still little difference between the stroke and healthy controls. However, as shown in FIG. 37C, when building a logistic regression classifier (with leave-out-out cross validation) to combine EV- associated alpha synuclein, phosphorylated alpha synuclein, and GFAP, a composite value of these measures created an enhanced PD classifier (AUC: 0.839), better than any individual measure. These results show that the level of EV-associated biomarkers can be normalized to the level of total EVs and subsequently combined by way of algorithmic correction to obtain an accurate Parkinson’s disease classifier. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 38A, 38B, and 38C show the results of detecting brain EV-associated (CD171 ) alpha- synuclein, phosphorylated alpha-synuclein, and respective EV quantity for Parkinson’s disease and healthy controls in a cohort of 14 symptomatic Parkinson’s Disease and 12 healthy controls (n=26). FIGs. 38A and 38B show plots of the signal for the healthy controls (circular points) and Parkinson’s disease controls (diamond points), while FIG. 38C shows plots of alpha synuclein aggregation prediction score for the healthy controls (filled circular points) and Parkinson’s disease controls (open diamond points) and of the ROC curve for determining the performance of the classifier model. FIG. 38A shows that each individual biomarker is not significantly different from the control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 38B, after regressing out the level of total brain EVs from the brain EV-associated biomarker signal there is still little difference between the stroke and healthy controls. However, as shown in FIG. 38C, when building a logistic regression classifier (with leave-out-out cross validation) to combine EV-associated alpha synuclein and phosphorylated alpha synuclein, a composite value of these measures created an enhanced PD classifier (AUC: 0.976), better than any individual measure. These results show that the level of brain EV-associated biomarkers can be normalized to the quantity of EVs within the corresponding EV subpopulation and subsequently combined by way of algorithmic correction to obtain an accurate PD classifier. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 39A, 39B, and 39C show the results of detecting neuron EV-associated (NrCAM) and astrocyte EV-associate (GLAST) alpha-synuclein, phosphorylated alpha-synuclein, and respective EV quantity for Parkinson’s disease and healthy controls in a cohort of 14 symptomatic Parkinson’s Disease and 12 healthy controls (n=26). FIGs. 39A and 39B show plots of the signal for the healthy controls (circular points) and Parkinson’s disease controls (diamond points), while FIG. 39C shows plots of alpha synuclein aggregation prediction score for the healthy controls (circular points) and Parkinson’s disease controls (diamond points) and of the ROC curve for determining the performance of the classifier model. FIG. 39A shows that each individual biomarker is not significantly different from the control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 39B, after regressing out the level of respective EV quantity from the neuron or astrocyte EV-associated biomarker signal there is still little difference between the stroke and healthy controls. However, as shown in FIG. 39C, when building a logistic regression classifier (with leave-out-out cross validation) to combine neuron and astrocyte EV-associated alpha synuclein and phosphorylated alpha synuclein, a composite value of these measures created an enhanced PD classifier (AUC: 0.964), better than any individual measure. These results show that the level of neuron and astrocyte EV-associated biomarkers can be normalized to the quantity of EVs within the corresponding EV subpopulation and subsequently combined by way of algorithmic correction to obtain an accurate PD classifier. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 40A, 40B, and 40C show the difference in performance between using total EV bound alpha synuclein, phosphorylated alpha synuclein, and GFAP (FIG. 40A, identical to FIG. 37C) and the brain EV subpopulation bound alpha synuclein and phosphorylated alpha synuclein (FIG. 40B, identical to FIG. 38C) or neuron and astrocyte EV subpopulation bound alpha synuclein and phosphorylated alpha synuclein (FIG. 40C, identical to FIG. 39C) in a cohort of 14 Parkinson’s Disease and 12 healthy controls (n=26). FIGs. 40A, 40B, and 40C show a plot on the left of the alpha synuclein aggregation prediction score for healthy (circular points) and Parkinson’s Disease (diamond points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. The performance of the total EV signature (AUC: 0.839) is not as good as the neural EV subpopulation signatures in FIGs. 40B and 40C (AUC: 0.976 or 0.964, respectively). Furthermore, the neural EV subpopulation signature only required forms of alpha synuclein as informative cargo, while the total EV signature required GFAP as well. Without being bound by any theory, it is hypothesized that the GFAP adds neural EV specificity to the total EV signature, enabling slightly better performance. This demonstrates that neural EV subpopulations in the blood can provide more information more reflective of the brain pathology compared to total EVs in the blood. These methods may extend to classifying other diseases resulting from alpha synuclein aggregation.

FIGs. 41 A, 41 B, and 41 C show the results of detecting EV-associated tau, phosphorylated tau (at amino acid positions 231 & 396), and total EV quantity for Tau-PET positive subjects and healthy controls. FIGs. 41 A and 41 B show plots of the signal for the healthy controls (filled circular points) and Tau-PET positive controls (open circular points), while FIG. 41C shows plots of Tau aggregation prediction score for the healthy controls (filled square points) and Tau-PET positive controls (filled circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 41 A shows that each individual biomarker is not significantly different in Tau-PET positive compared to control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 41 B, after regressing out the level of total EVs from the EV-associated biomarker signal there is still little difference between the Tau-PET positive and healthy controls. FIG. 41 C shows that building a logistic regression classifier (with leave-out-out cross validation) to combine EV-associated tau and phosphorylated tau, there was still little ability to classify Tau-PET positivity (AUC: 0.590).

FIGs. 42A, 42B, and 42C show the results of detecting brain (CD171 ) and astrocyte (GLAST) EV- associated tau, phosphorylated tau (at amino acid positions 231 & 396), and total EV quantity for Tau-PET positive subjects and healthy controls. FIGs. 42A and 42B show plots of the signal for the healthy controls (filled circular points) and Tau-PET positive controls (open circular points), while FIG. 42C shows plots of Tau aggregation prediction score for the healthy controls (filled square points) and Tau-PET positive controls (filled circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 42A shows that each individual biomarker is not significantly different in Tau-PET positive compared to control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 42B, after regressing out the respective level of brain and astrocyte EV quantity from the EV-associated biomarker signal there is better separation between the Tau-PET positive and healthy controls. However, as shown in FIG. 42C, when building a logistic regression classifier (with leave-out-out cross validation) to combine EV-associated tau and phosphorylated tau, a composite value of these measures created an enhanced Tau-PET positivity classifier (AUC: 0.835), better than any individual measure. These results show that the level of brain and astrocyte EV-associated biomarkers can be normalized to the quantity of EVs within the corresponding EV subpopulation and subsequently combined by way of algorithmic correction to obtain an accurate brain Tau-PET positivity classifier. These methods may extend to classifying other diseases resulting from tau aggregation.

FIGs. 43A, 43B, and 43C show the results of detecting neuron (NrCAM) EV-associated tau, phosphorylated tau (at amino acid positions 231 & 396), and total EV quantity for Tau-PET positive subjects and healthy controls. FIGs. 43A and 43B show plots of the signal for the healthy controls (filled circular points) and Tau-PET positive controls (open circular points), while FIG. 43C shows plots of Tau aggregation prediction score for the healthy controls (filled square points) and Tau-PET positive controls (filled circular points) and of the ROC curve for determining the performance of the classifier model. FIG. 43A shows that each individual biomarker is not significantly different in Tau-PET positive compared to control when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 43B, after regressing out the level of neuron EV quantity from the EV-associated biomarker signal there is better separation between the Tau-PET positive and healthy controls. However, as shown in FIG. 43C, when building a logistic regression classifier (with leave-out-out cross validation) to combine EV-associated tau and phosphorylated tau, a composite value of these measures created an enhanced Tau-PET positivity classifier (AUC: 0.917), better than any individual measure. These results show that the level of neuron EV- associated biomarkers can be normalized to the quantity of EV subpopulation and subsequently combined by way of algorithmic correction to obtain an accurate brain Tau-PET positivity classifier. These methods may extend to classifying other diseases resulting from tau aggregation.

FIGs. 44A, 44B, and 44C show the difference in performance between using total EV bound tau and phosphorylated tau (FIG. 44A, identical to FIG. 41 C) and the brain and astrocyte EV subpopulation bound tau and phosphorylated tau (FIG. 44B, identical to FIG. 42C) or neuron EV subpopulation bound tau and phosphorylated tau (FIG. 44C, identical to FIG. 43C) in a cohort of 16 Tau-PET positive subjects and 18 healthy controls (n=34). FIGs. 44A, 44B, and 44C show a plot on the left of the tau aggregation prediction score for healthy (square points) and Tau-PET positive (circular points) controls and a plot on the right of the ROC curve for determining the performance of the classifier model. The performance of the total EV signature (AUC: 0.593) is not as good as the neural EV subpopulation signatures in FIGS. 44B and 44C (AUC: 0.835 or 0.917, respectively). This demonstrates that neural EV subpopulations in the blood can provide more information more reflective of the brain pathology compared to total EVs in the blood. These methods may extend to classifying other diseases resulting from tau aggregation.

FIGs. 45A and 45B shows the performance of the neuron (NrCAM) EV-associated tau, phosphorylated tau (at amino acid positions 231 & 396 corresponding to pTau231 and pTau 396, respectively), and total EV quantity for the original cohort of 16 Tau-PET positive subjects, 18 healthy controls, and an expansion of addition 14 healthy controls, 2 Tau-PET positive, Amyloid-PET negative, and 4 Tau-PET negative, Amyloid-PET positive subjects. FIG. 45A shows a plot of the tau aggregation prediction score for the healthy controls of the first cohort (square points), Tau-PET positive and amyloid positive controls (open circular points), healthy controls of the second cohort (filled circular points), Tau-PET positive and amyloid-PET negative controls (filled triangular points), and Tau-PET negative and amyloid-PET negative controls (open triangular points). FIG. 45B shows a plot of the ROC curve for determining the performance of the classifier model. The classifier trained on the original cohort was able to classify the samples in the expanded cohort accurately (FIG. 45A). Based on a cutoff that gave 91% accuracy, 13/14 healthy controls were correctly classified, 2/2 Tau-PET positive, Amyloid-PET negative samples were correctly classified, and 4/4 Tau-PET negative, Amyloid-PET positive samples were correctly classified. This shows that neural EV-associated forms of tau cargo are specific to tau aggregation in the brain and not amyloid aggregation. FIG. 45B shows the performance of the logistic regression classifier (with leave-out-out cross validation) on the original 16 Tau-PET positive subjects and all 32 healthy samples in the expanded cohort. Its performance (AUC: 0.936) is comparable to the previous model (FIG. 44C, AUC: 0.917), showing that the classifier is specific to tau pathology and not to sample collection or site differences. These methods may extend to classifying other diseases resulting from tau aggregation.

FIGs. 46A and 46B is a schematic showing expressing recombinant proteins (FIG. 46A) and methods of isolating extracellular vesicles (FIG. 46B).

FIGs. 47A, 47B, and 47C show the results of detecting EV-associated transactive response DNA binding protein of 43 kDa (TDP43), EV-associated phosphorylated TDP43, and total EV quantity for subjects having amyotrophic lateral sclerosis (ALS) (13 subjects) and healthy controls (23 subjects). FIGs. 47A, 47B, and 47C show plots of the signal and TDP43 aggregation prediction score for the healthy controls (filled circular points) and ALS controls (open circular points), along with a plot of the ROC curve for determining the performance of the classifier model in FIG. 47C. FIG. 47A shows that each individual biomarker is not significantly different in ALS subjects compared to the healthy controls when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 47B, after regressing out the level of total EV quantity from the EV-associated biomarker signal, each biomarker was not better able to classify the two populations. As shown in FIG. 47C, when building a logistic regression classifier (with leave- out-out cross validation) to combine the corrected EV-associated TDP43 and phosphorylated TDP43, a combined value of these composite values did not create an enhanced ALS classifier (AUC: 0.599). FIGs. 48A, 48B, and 48C show the results of detecting EV-associated transactive response DNA binding protein of 43 kDa (TDP43), EV-associated phosphorylated TDP43, total neuron EV quantity (characterized by NrCAM), and total neural lineage EV quantity (characterized by CD56) for subjects having amyotrophic lateral sclerosis (ALS) (13 subjects) and healthy controls (23 subjects). FIGs. 48A, 48B, and 48C show plots of the signal and TDP43 aggregation prediction score for the healthy controls (filled circular points) and ALS controls (open circular points), along with a plot of the ROC curve for determining the performance of the classifier model in FIG. 48C. FIG. 48A shows that each individual biomarker is not significantly different in ALS subjects compared to the healthy controls when using the absolute concentration of the biomarker per unit volume of the blood. As shown in FIG. 48B, after regressing out the level of corresponding total neuron EV quantity or total neural EV quantity from the EV-associated biomarker signal, each biomarker was not better able to classify the two populations. However, as shown in FIG. 48C, when building a logistic regression classifier (with leave-out-out cross validation) to combine the corrected EV-associated TDP43 and phosphorylated TDP43, a composite value of these measures created an enhanced ALS classifier (AUC: 0.893) as compared to the total EV-corrected biomarkers algorithmically combined in FIG. 47C (AUC: 0.599). These methods may extend to classifying other diseases resulting from TDP43 aggregation.

DETAILED DESCRIPTION

The application provides useful methods and compositions related to technology that analyzes the biomarkers bound to extracellular vesicles. EVs used in the assays can be obtained in several ways. The methods also relate to diagnosing a subject with a disease on the basis of their levels of EV-associated biomarkers, along with treating the subject based on their diagnosis.

Extracellular Vesicles (EVs)

Extracellular vesicles (EVs) serve as carriers of vital biomolecules, including proteins, nucleic acids, and lipids. These biomolecules can be adorned with specific markers, a characteristic that may arise through two distinct processes: active secretion by the parent cells and adsorption from the neighboring extracellular environment.

In the first scenario, cells spontaneously release EVs as part of their regular physiological functions. During this release, the EVs incorporate biomarkers that reflect the molecular identity and status of the parent cell. These markers, often proteins such as receptors or signaling molecules, act as cellular signatures, offering insights into the tissue source and its condition.

Alternatively, EVs can acquire biomarkers through adsorption in the extracellular space. This process involves the association of molecules from the local environment onto the surface of EVs. The adsorbed biomarkers could originate from neighboring cells or the surrounding microenvironment, providing EVs with a dynamic snapshot that mirrors the molecular complexity of both the parent cell and its immediate surroundings.

This means that different types of EVs will bind differentially to the same biomarker depending on its biogenesis origin and/or cellular source local microenvironment. Biogenesis, secretion, intercellular interactions, and other aspects of the cellular biology of EVs are described in, e.g., Colombo, M., Raposo, G. & Thery, C. Biogenesis, secretion, and intercellular interactions of exosomes and other extracellular vesicles. Annu Rev Cell Dev Biol 30, 255-289 (2014); Van Niel, G., D’Angelo, G. & Raposo, G. Shedding light on the cell biology of extracellular vesicles. Nat Rev Mol Cell Biol 19, 213-228 (2018); Choi, D. S., Kim, D. K., Kim, Y. K. & Gho, Y. S. Proteomics of extracellular vesicles: Exosomes and ectosomes. Mass Spectrom Rev 34, 474-490 (2015); and Maas, S. L. N., Breakefield, X. O. & Weaver, A. M. Extracellular Vesicles: Unique Intercellular Delivery Vehicles. Trends Cell Biol 27, 172-188 (2017). all of which are herein incorporated by reference in their entirety.

Examples of transfer of RNA and proteins between cells by way of EVs are described in, e.g., Valadi, H. et al., Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells. Nat Cell Biol 9, 654-659 (2007); and Skog, J. et al., Glioblastoma microvesicles transport RNA and proteins that promote tumor growth and provide diagnostic biomarkers: Nat Cell Biol 10, 1470- 1476 (2008), both herein incorporated by reference in their entirety.

In some embodiments, magnetic-bead based methods may transduce EV-associated biomarkers into magnetic signals (see, e.g., Wang, Z. et al., Dual-selective magnetic analysis of extracellular vesicle glycans. Matter 2, 150-166 (2020), herein incorporated by reference in its entirety).

In some embodiments, EVs can be isolated from in vitro cell culture, e.g., tumor cell lines, as described below. In some embodiments, EVs can be isolated from a bodily fluid {e.g., a blood sample) or a sample prepared from a tissue {e.g., a tumor biopsy) from a patient by differential centrifugation. Methods for isolating EVs typically employ a series of centrifugation steps with increasing centrifugal force to separate the extracellular vesicles from cells, cell debris, and other larger cellular particles. In one embodiment, the blood sample can first be centrifuged at 10,000 g to remove debris and/or apoptotic bodies and subsequently at 100,000 g to precipitate EVs. The extracellular vesicles are then collected, washed, and resuspended in a suitable buffer, e.g., PBS. The prepared extracellular vesicles can be stored at -80 ^eC for future use. In some embodiments, extracellular vesicles can be isolated using size-exclusion chromatography. Suitable size exclusion chromatography, such as sepharose 2B columns, is commercially available from Sigma Aldrich (St. Louis, MO). The columns are prepared according to the manufacturer’s instructions. In some embodiments, magnetic bead-based methods may isolate EVs from the sample. Such methods, such as Dynabeads, may be commercially available from Thermo Fisher Scientific, Waltham, MA. Other methods may be suitable for EV isolation, including, but not limited to, microfluidic devices, precipitation, immunoaffinity isolation, or density gradient ultracentrifugation.

FIGs. 46A and 46B illustrate an exemplary procedure for recombinant protein extraction and isolating EVs from a cell culture, respectively. A protein of interest may be synthesized by first identifying a gene sequence of interest and extracting the information from a software database, such as Uniprot. The codons may be optimized for expression, and sequence of interest may be synthesized in an expression vector for cloning. The vector is then transfected into the host cells for expression, followed by isolation of the protein and purification (FIG. 46A). EVs may be collected from culture media that contain the EVs released by the host cells transfected to express a gene of interest through a series of centrifugation steps as described in FIG. 46B. In some embodiments, the concentration and size distribution of the EVs can be analyzed using commercially available devices, for example, the nanoparticle tracking analysis (NTA) system (Nanosight, NS300). In one embodiment, extracellular vesicles can be confirmed using Western blots to detect EV proteins. In some embodiments, the size and morphology of the extracellular vesicles can be confirmed using methods such as flow cytometry and transmission electron microscopy.

Sample

A sample used in the methods and composition disclosed herein can be any sample comprising or being tested for the presence of a biomarker. Samples include but are not limited to, samples derived from or containing cells, organisms (bacteria, viruses), lysed cells or organisms, cellular extracts, nuclear extracts, components of cells or organisms, extracellular fluid, media in which cells or organisms are cultured in vitro, blood, plasma, serum, gastrointestinal secretions, urine, ascites, homogenates of tissues or tumors, synovial fluid, feces, saliva, sputum, cyst fluid, amniotic fluid, cerebrospinal fluid, peritoneal fluid, lung lavage fluid, semen, lymphatic fluid, tears, pleural fluid, nipple aspirates, breast milk, external sections of the skin, respiratory, intestinal, and genitourinary tracts, and prostatic fluid. A sample can be a viral or bacterial sample obtained from an environmental source, such as a body of polluted water, an air sample, a soil sample, or a food industry sample. A sample can be a biological sample, which refers to the fact that it is derived or obtained from a living organism. The organism can be in vivo {e.g., a whole organism) or in vitro {e.g., cells or organs grown in culture). A sample can be a biological sample, which refers to a cell or population of cells or a quantity of tissue or fluid from a subject. Most often, a sample has been removed from a subject. Still, the term “biological sample” can also refer to cells or tissue analyzed in vivo, e.g., without removal from the subject. Often, a “biological sample” will contain cells from a subject. Still, the term can also refer to non-cellular biological material, such as non-cellular fractions of blood, saliva, or urine. The biological sample may be from a resection, bronchoscopic biopsy, core needle biopsy of a primary, secondary, or metastatic tumor, or a cell block from pleural fluid. In addition, fine needle aspirate biological samples are also useful. In one embodiment, a biological sample is primary ascites cells. Biological samples also include explants and primary and/or transformed cell cultures derived from patient tissues. A biological sample can be provided by removing a sample of cells from the subject. Still, it can also be accomplished by using previously isolated cells or cellular extracts {e.g., isolated by another person, at another time, and/or for another purpose). Archival tissues, such as those having treatment or outcome history, may also be used. Biological samples include, but are not limited to, tissue biopsies, scrapes {e.g., buccal scrapes), whole blood, plasma, serum, urine, saliva, cell culture, or cerebrospinal fluid. The samples analyzed by the compositions and methods described herein may have been processed for purification or enrichment of exosomes contained therein. In one embodiment, the sample is blood.

Sample Diluent

Diluents commonly used for reducing non-specific binding in assays include animal serum. Unless treated, the animal serum can contain exogenous EVs, which are shown to disrupt the signal from the EV- associated biomarker in the present invention. The present invention includes the use of a diluent that is substantially free of exogenous EVs. In some embodiments, the sample is diluted with a diluent. In some embodiments the sample is diluted with a diluent after the step of binding the EVs in the assay by way of an affinity agent. In some embodiments, the diluent includes water, protein, buffer, salt, polymer, preservative, and/or detergent (i.e., surfactant). In some embodiments, the components of the diluent are purified to remove any EVs, wherein the EVs are human-derived and/or non-human derived (e.g., exogenous EVs, e.g., cell-line derived EVs or animal-derived EVs). In some embodiments, the protein of the diluent competitively inhibits non-specific binding of biomolecules within the sample. Such biomolecules within the sample may bind generally to all antibodies, generating a false positive signal. The protein of the diluent may be an IgG, IgM, and/or IgA. The antibody may originate from, but not limited to, a goat, hamster, rabbit, hamster, rat, and/or human. In some embodiments, the buffer of the diluent includes a pH within a physiological range. In some embodiments, the physiological range of the pH is from about 7.0 to about 8.0. In some embodiments, the buffer of the diluent preserves EV integrity and protein folding. In some embodiments, the salt of the diluent maintains the osmotic potential and the net charge of the sample within the physiological ranges to preserve EV integrity and protein folding. In some embodiments, the polymer of the diluent helps to increase the interactions between the target and binding agent by increasing the viscosity of the sample and inducing macromolecular crowding. In some embodiments, the polymer increases the preferential binding of the first binding agent to the source marker to separate the EV subpopulation from the mixture of EVs in the sample to form an isolated sample. In some embodiments, the polymer increases the preferential binding of the second binding agent to the EV-associated form of a disease biomarker. The polymer of the diluent may include polyethylene glycol (PEG), polyvinylpyrrolidone (PVP), dextran, mannitol, betaine, mannitol, sorbitol, xylitol, or other commonly known and used stabilizers. In some embodiments, the preservative helps to maintain the long-term sterility of the sample. The preservative of the diluent may include sodium azide, ProCiin™, thimerosal, and/or sodium benzoate. In some embodiments, the detergent (i.e., surfactant) of the diluent helps to minimize non-specific binding of biomolecules within the sample to the surface of the assay. The detergent of the diluent may include Tween® 20, Tween® 40, Tween® 60, Tween® 80, BRIJ® 020, Triton™ X-100, Triton™ X-114, and/or IGEPAL® (see, e.g., FIGs. 14A and 14B). The stability of EV-associated biomarkers after incubation in the detergent, Triton™ X-100, is shown in FIGs. 12A and 12B. In some embodiments, the detergent of the diluent permeabilizes the membrane of the EV. Without being bound by any theory, the inherent resilience of EVs to common lysis conditions results in a specific phenomenon where the common procedure of sequentially isolating EVs, exposing them to a lysis solution, and measuring the lysate/supernatant results in a much lower signal than directly measuring the isolated EVs and their associated cargo (see, e.g., FIGs. 11 , 13, and 14B). By permeabilizing the EV membrane, the signal is further improved due to affinity agents gaining access the internal cargo of the EV along with greater access to binding sites of the external EV cargo (e.g., EV-associated biomarkers and EV markers).

Permeabilizing Agents

The sample preparation methods of the invention can include exposing an EV to be detected (e.g., an EV bound to a surface) to a permeabilizing agent prior to a detection step (e.g., with sample dilution and/or following isolation of a target EV on a surface, such as in a sandwich assay). The mild nonionic surfactant can be a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate surfactant. Polysorbate surfactants

Polysorbate surfactants Polysorbate surfactants can be used as EV permeabilizing nonionic surfactants of the invention. Polysorbate surfactants are oily liquids derived from pegylated sorbitan esterified with fatty acids. Common brand names for Polysorbates include Alkest, Canarcel and Tween. Polysorbate surfactants include, without limitation, polyoxyethylene 20 sorbitan monolaurate (TWEEN® 20), polyoxyethylene (4) sorbitan monolaurate (TWEEN® 21 ), polyoxyethylene 20 sorbitan monopalmitate (TWEEN® 40), polyoxyethylene 20 sorbitan monostearate (TWEEN® 60); and polyoxyethylene 20 sorbitan monooleate (TWEEN® 80). A permeabilizing amount of polysorbate surfactant can be, e.g., from about 0.01 to 0.75% (w/w) of a rinse solution (e.g., about 0.02 ± 0.01% (w/w), 0.05 ± 0.025% (w/w), 0.1 ± 0.05% (w/w), 0.2 ± 0.1% (w/w), 0.35 ± 0.1% (w/w), or 0.5 ± 0.25% (w/w) polysorbate surfactant. For example, the rinse solution can contain about 0.05 to 0.25% (w/w) or about 0.1% (w/w) polysorbate surfactant.

Polyethylene glycol alkyl ethers

Ethers of polyethylene glycol and alkyl alcohols can be used as EV permeabilizing nonionic surfactants of the invention. Preferred polyethylene glycol alkyl ethers include Laureth 9, Laureth 12 and Laureth 20. Other polyethylene glycol alkyl ethers include, without limitation, PEG-2 oleyl ether, oleth-2 (Brij® 92/93, Atlas/ICI); PEG-3 oleyl ether, oleth-3 (Volpo 3, Croda); PEG-5 oleyl ether, oleth-5 (Volpo 5, Croda); PEG-10 oleyl ether, oleth-10 (Volpo 10, Croda, Brij® 96/97 12, Atlas/ICI); PEG-20 oleyl ether, oleth-20 (Volpo 20, Croda, Brij® 98/99 15, Atlas/ICI); PEG-4 lauryl ether, laureth-4 (Brij® 30, Atlas/ICI); PEG-9 lauryl ether; PEG-23 lauryl ether, laureth-23 (Brij™ 35, Atlas/ICI); PEG-2 cetyl ether (Brij® 52, ICI); PEG-10 cetyl ether (Brij® 56, ICI); PEG-20 cetyl ether (Brij™ 58, ICI); PEG-2 stearyl ether (Brij™ 72, ICI); PEG-10 stearyl ether (Brij® 76, ICI); Polyoxyethylene (20) oleyl ether (Brij® 020); PEG-20 stearyl ether (Brij™ 78, ICI); and PEG- 100 stearyl ether (Brij® 700, ICI). A permeabilizing amount of polyethylene glycol alkyl ethercan be, e.g., from about 0.01 to 2.5% (w/w) of a rinse solution (e.g., about 0.02 ± 0.01% (w/w), 0.05 ± 0.025% (w/w), 0.1 ± 0.05% (w/w), 0.2 ± 0.1% (w/w), 0.35 ± 0.1% (w/w), 0.5 ± 0.25% (w/w), 0.75 ± 0.25% (w/w), 1 .0 ± 0.25% (w/w), 1 .5 ± 0.25% (w/w), 1 .75 ± 0.25% (w/w), or 2.25 ± 0.25% (w/w) polyethylene glycol alkyl ether. For example, the rinse solution can contain about 0.5 to 1 .25% (w/w) or about 0.75% (w/w) polyethylene glycol alkyl ether.

Alkylphenol ethoxylates

Alkylphenol ethoxylates can be used as EV permeabilizing nonionic surfactants of the invention. Preferred alkylphenol ethoxylates include polyethylene glycol tert-octylphenyl ether (Triton™ X-114), and 2- [4-(2,4,4-trimethylpentan-2-yl)phenoxy]ethanol (Triton™ X-100, Igepal® CA-210, octoxynol-9) A permeabilizing amount of alkylphenol ethoxylate can be, e.g., from about 0.1 to 2.5% (w/w) of a rinse solution (e.g., about 0.1 ± 0.05% (w/w), 0.2 ± 0.1% (w/w), 0.35 ± 0.1% (w/w), 0.5 ± 0.25% (w/w), 0.75 ± 0.25% (w/w), 1 .0 ± 0.25% (w/w), 1 .5 ± 0.25% (w/w), 1 .75 ± 0.25% (w/w), or 2.25 ± 0.25% (w/w) alkylphenol ethoxylate. For example, the rinse solution can contain about 0.25 to 1 .25% (w/w) or about 0.75% (w/w) alkylphenol ethoxylate. For example, the method can include (a) providing a sample including a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface including the EV-associated biomarker cargo; (b) diluting the sample with a diluent comprising a protein, wherein the diluent is substantially free of exogenous EVs; (c) following step (b), capturing the EV subpopulation having a surface including the EV-associated biomarker cargo on a surface; (d) washing the surface with a mild nonionic surfactant to produce permeabilized EV-associated biomarker cargo; (e) using a binding agent that preferentially binds to the permeabilized EV-associated biomarker cargo on the surface; and (f) measuring the presence of, or level of, the permeabilized EV-associated biomarker cargo.

Biomarkers for Detecting a Diseased State

Disclosed herein are methods and compositions for detecting biomarkers indicating a diseased state in a subject. Nonlimiting examples of biomarkers that can be detected using the methods disclosed herein include whether the biomarkers may be soluble or EV-associated. Nonlimiting examples include CD9, CD63, CD81 , CD56, CD171 , NrCAM, GLAST, EAAT1 , SYP, Alpha Synuclein (aSyn), phosphorylated Alpha Synuclein (paSyn), Tau, phosphorylated Tau (pTau, e.g., pTau231 and pTau396), Amyloid beta (Ap), Amyloid Beta 42 peptide (Ap42), TDP43, phosphorylated TDP43 (pTDP43), GFAP, NfL, SWOB, NSE, IL-6, IL-1 beta, and TNF-alpha. GLAST and EAAT1 refer to the same biomarker and are, thus, used interchangeably when referring to the biomarker. SWOB and its implications in the regulation of a variety of cellular activities is described in, e.g., Donato R. intracellular and extracellular roles of S100 proteins. Microscopy Research and Technique 60:540-551 (2003); Adami C, Sorci G, Blasi E, Agneletti AL, Bistoni F, Donato R. SWOB expression in and effects on microglia. Glia 33:131 -142 (2001 ); and Donato R, S100: A multigenic family of calcium-modulated proteins of the EF-hand type with intracellular and extracellular functional roles. International journal of biochemistry & cell biology 33:637-668 (2001 ), all of which are herein incorporated by reference in their entirety.

In some embodiments, the biomarker is a disease process biomarker for stroke. In some embodiments, the stroke biomarker is NSE or SWOB. For example, NSE and SWOB as predicters for stroke have been described in, e.g., Kaca-Orynska M, Tomasiuk R, Friedman A. Neuron-specific eno-lase and SWOB protein as predictors of outcome in ischaemic stroke. Neurol Neurochir Pol. 44:459-463 (2010), herein incorporated by reference in its entirety.

In some embodiments, the biomarker is a disease process biomarker for cancer. In some embodiments, the cancer biomarker is NSE or TNF-alpha.

In some embodiments, the biomarker is a disease process biomarker for neurological diseases. For example, a disease process biomarker may include CD56, CD171 , NrCAM, EAAT1 , SYP, Tau, phosphorylated Tau, Amyloid beta, alpha-synuclein, phosphorylated alpha-synuclein, TDP43, phosphorylated TDP43, GFAP, NfL, SWOB, NSA, IL-6, 11-1 beta, and TNF-alpha. In some embodiemnts, the biomarker is a disease process biomarker for ALS. For examples, the disease process biomarker for ALS may include TDP43 and phosphorylated TDP 43. SWOB has been studied as a biological marker diagnostic of severe head injury, brain damage, and neurodegeneration (see, e.g., Raabe A, Kopetsch O, Woszczyk A, Lang J, Gerlach R, Zimmermann M, Seifert V. serum S-100B protein as a molecular marker in severe traumatic brain injury. Restorative Neurology and Neuroscience 21 :159-169 (2003); and Raabe A, Grolms C, Sorge O, Zimmermann M. Serum S-100B protein in severe head injury. Neurosurgery 45:477-483 (1999); Rothermundt M, Peters M, Preehn JH, Arolt O. SWOB in brain damage and neurodegeneration. Microsc res tech 60:614-632 (2003), all of which are herein incorporated by reference in their entirety). Alpha-synuclein fibrils have been identified as the major component of Lewy bodies, which are characteristic of Parkinson’s disease (see, e.g., Spillantini MG, Schmidt ML, Lee VM, Trojanowski JQ, Jakes R, Goedert M. Alpha synuclein in Lewy bodies. Nature. 388:839-40 (1997); herein incorporated by reference in its entirety).

In some embodiments, the biomarker is a disease process biomarker for inflammation or autoimmune disease. For example, the inflammation or autoimmune disease biomarker may be CD56, GFAP, IL-6, IL-1 beta, or TNF-alpha.

A biomarker disclosed herein may be in soluble form or an EV-associated form. In some embodiments, the soluble form of the biomarker may be present in the patient sample at the same level as the EV-associated biomarker. In some embodiments, the soluble biomarker may be at a lower level than the EV-associated biomarker.

In some embodiments, the biomarker binds to the EVs via covalent binding, non-covalent binding, direct binding via modifications or lipid-protein interactions, or indirect binding via protein-DNA/RNA interactions or protein-protein interactions.

In some embodiments, when a soluble biomarker binds to an EV to become an EV-associated biomarker, the biomarker protein undergoes a conformational change. Thus, the EV-associated biomarker may possess a unique conformation different from the soluble form of the same biomarker. In some cases, the conformational change results in epitopes normally embedded inside the molecule exposed and available for binding by the binding agents. These epitopes are referred to as conformational epitopes in this disclosure.

Biomarkers for Isolating EV Subpopulations

Disclosed herein are methods for isolating EV subpopulations, EVs which pertain to a specific cellular origin, from a mixture of EVs within a sample. Nonlimiting examples of cellular origins include brain cells, endothelial cells, epithelial cells, glial cells, astrocyte cells, pericytes, vascular endothelial cells, or neuronal cells. Also disclosed herein are methods for detecting EV-associated biomarkers from an EV subpopulation indicating a diseased state in a subject. Nonlimiting examples of EV markers indicative of an EV subpopulation include MOG, SOX10, OLIG1 , IBA1 , TMEM119, TREM2, EAAT1 , GLAST, EAAT2, CD49f, L1 CAM, CD171 , NCAM, CD56, NrCAM, NRXN3, SYP, SYN1 , PDGFR-beta, CSPG4, CD13, SLC2A1 or CD31.

In some embodiments, the EV subpopulation biomarker corresponds to an oligodendrocyte cellular origin. For example, the oligodendrocytes EV subpopulation biomarker may be MOG, SOX10, or OLIG1.

In some embodiments, the EV subpopulation biomarker corresponds to a microglia cellular origin. For example, the microglia EV subpopulation biomarker may be IBA1 , TMEM119, or TREM2.

In some embodiments, the EV subpopulation biomarker corresponds to an astrocyte cellular origin. For example, the microglia EV subpopulation biomarker may be EAAT1 , GLAST, EAAT2, or CD49f. In some embodiments, the EV subpopulation biomarker corresponds to a neuronal cellular origin. For example, the neuronal EV subpopulation biomarker may be L1 CAM, CD171 , NCAM, CD56, NrCAM, or NRXN3.

In some embodiments, the EV subpopulation biomarker corresponds to a synapse cellular origin. For example, the synapse EV subpopulation biomarker may be SYP or SYN1 .

In some embodiments, the EV subpopulation biomarker corresponds to a pericyte cellular origin. For example, the pericyte EV subpopulation biomarker may be PDGFR-beta, CSPG4, or CD13.

In some embodiments, the EV subpopulation biomarker corresponds to a vascular endothelial cellular origin. For example, the vascular endothelial EV subpopulation biomarker may be SLC2A1 or CD31 .

Biomarkers expressed on the surface of the EVs (EV markers) are also disclosed. Nonlimiting examples of the EV markers include CD9, CD56, CD81 , CD63, TSG101 , CD171 , NrCAM, GLAST, EAAT1 , SYP, ALIX, HSP70, MHC 1 , MHC 2, MOG, SOX10, OLIG1 , IBA1 , TMEM119, TREM2, EAAT2, CD49f, L1 CAM, NCAM, NRXN3, SYN1 , PDGFR beta, CSPG4, CD13, SLC2A1 , and CD31. Pan biomarkers (e.g., universal biomarkers) expressed on the surface of the EVs (EV markers) include, but are not limited to, CD9, CD63, CD81 , CD56, CD171 , NrCAM, EAAT1 , SYP, ALIX, HSP70, MHC 1 , and MHC 2.

Binding Agents

Aspects of the invention involve binding agents (affinity agents) in detecting a biomarker’s presence. In some embodiments, the binding agents used in the methods disclosed herein can specifically bind to the biomarker but may display differential affinity to different forms of the same biomarker. Further disclosed below, multiple types of binding agents can be used in the detection methods.

EVbA

As used herein, an EVbA refers to a binding agent that preferentially or exclusively binds to an EV- associated form of the biomarker rather than a soluble form of the biomarker. In some cases, the binding agent binds specifically to an EV-associated biomarker at newly exposed or configured epitopes. These epitopes are referred to as conformational epitopes in this disclosure. In some embodiments, the EVbA binds to one or more conformational epitopes of the biomarker.

An EVbA can be readily identified by comparing binding affinities between the EV-associated and soluble biomarkers, for example, by measuring the affinity of different antibodies against soluble recombinant protein versus EV-associated forms of the biomarkers isolated from cell lines or a subject (e.g., human blood or plasma), using common methods such as ELISA or Surface Plasmon Resonance. Illustration of methods for producing recombinant protein and EV-associated forms are depicted in FIGs. 46A and 46B, respectfully. An EVbA can be identified if it shows preferential binding to the EV-associated form of the biomarker as compared to its soluble form.

Several EVbAs have been identified and used to detect EV-associated biomarkers. These EVbAs were able to detect the biomarkers such as CD9, CD56, CD81 , CD63, TSG101 , CD171 , NrCAM, GLAST, EAAT1 , SYP, ALIX, HSP70, MHC 1 , MHC 2, MOG, SOX10, OLIG1 , IBA1 , TMEM119, TREM2, EAAT2, CD49f, L1 CAM, NCAM, NRXN3, SYN1 , PDGFR beta, CSPG4, CD13, SLC2A1 , CD31 , Tau, phosphorylated Tau, Amyloid beta, Amyloid beta 42 peptide, alpha-synuclein, phosphorylated alpha-synuclein, TDP43, phosphorylated TDP43, GFAP, NfL, SWOB, NSA, IL-6, 11-1 beta, and TNF-alpha.

SSA

As used herein, an SSA refers to a binding agent that preferentially or exclusively binds to a soluble form of the biomarker rather than an EV-associated form of the biomarker. A SSA can be readily identified by comparing 1 ) binding to the soluble form of biomarker to a control protein and 2) comparing binding to the EV-associated form of the biomarker to a control protein. A SSA is identified if it shows 1 ) a significantly greater binding to the soluble form of a biomarker than a control protein and 2) no significantly greater binding to the EV-associated form of a biomarker than a control protein.

UA

As used herein, a UA refers to a binding agent that binds to both a soluble form of the biomarker and its associated EV-associated form. A UA can be readily identified by comparing 1 ) binding to the soluble form of biomarker to a control protein and 2) comparing binding to the EV-associated form of the biomarker to a control protein. A UA is identified if it shows 1 ) a significantly greater binding to the soluble form of a biomarker than a control protein and 2) a significantly greater binding to the EV-associated form of a biomarker than a control protein.

EVA

As used herein, an EVA refers to a binding agent that binds to pan (or, e.g., canonical, integral, or universal) EV-associated biomarkers (EV markers). Nonlimiting examples of integral EV markers include CD9, CD63, CD81 , CD56, CD171 , NrCAM, EAAT1 , SYP, ALIX, HSP70, MHC 1 , and MHC 2. In some cases, detecting a pan EV marker is used to identify and quantify EVs, which can be used to normalize the level of EV-associated biomarkers relative to the level of total EVs. In some cases, a pan EV marker can also serve as a disease process marker; for example, CD56 is a pan EV marker as well as a disease process biomarker for inflammatory or autoimmune diseases.

Types of binding agents

A binding agent disclosed herein, for example, an EVbA, an SSA, a UA, or an EVA, may be an antibody, antibody fragment, an aptamer, a peptide, a specific binding protein, a nucleic acid, a small molecule, or various synthetic agents.

In some embodiments, an antibody used as a binding agent of this disclosure includes, but is not limited to, synthetic antibodies, monoclonal antibodies, recombinantly produced antibodies, multispecific antibodies (including bi-specific antibodies), human antibodies, humanized antibodies, chimeric antibodies, single-chain Fvs (scFv), Fab fragments, F(ab') fragments, disulfide-linked Fvs (sdFv) (including bi-specific sdFvs), and anti-idiotypic (anti-ld) antibodies, and epitope-binding fragments of any of the above. The antibodies provided herein may be monospecific, bispecific, trispecific, or of greater multi-specificity. Multispecific antibodies may be specific for different epitopes of a biomarker disclosed herein or may be specific for both a biomarker as well as for a heterologous epitope.

In some embodiments, nucleic acid aptamers or “aptamers” used as binding agents are nucleic acid species that have been engineered through repeated rounds of in vitro selection or, equivalently, SELEX (systematic evolution of ligands by exponential enrichment) to bind to various molecular targets, such as the biomarkers described herein. For example, aptamers with an affinity for a target biomarker can be selected from an extensive oligonucleotide library through SELEX, an iterative process in which non-binding aptamers are discarded, and aptamers binding to the proposed target are expanded. Initial positive selection rounds are sometimes followed by negative selection. This negative selection improves the selectivity of the resulting aptamer candidates. In this process, the target is immobilized to an affinity column. The aptamer library is applied and allowed to bind. Weak binders are washed away, and bound aptamers are eluted and amplified using PCR. Then, the pool of amplified aptamers is reapplied to the targets. The process is repeated multiple times under increasing stringency until aptamers of the desired selectivity and affinity are obtained. See, e.g., Jayasena et al., Clinical Chemistry 45:1628-1650, 1999, herein incorporated by reference in its entirety. Peptide aptamer can be selected using different systems, including the yeast two- hybrid system. Peptide aptamers can also be selected from combinatorial peptide libraries constructed by phage display and other surface display technologies such as mRNA display, ribosome display, bacterial display, and yeast display. These experimental procedures are also known as biopanning. See, e.g., Reverdatto et al., 2015, Curr. Top. Med. Chem. 15:1082 1101 , herein incorporated by reference in its entirety.

In some embodiments, the methods described herein may include a first binding agent wherein the first binding i) preferentially binds or exclusively binds to an EV-associated form of the biomarker or ii) binds to both the EV-associated form of the biomarker and a soluble form of the biomarker. In some embodiments, the methods described herein may further include a second binding agent that binds to i) an EV-associated biomarker, ii) to both an EV-associated biomarker and a soluble form of the biomarker, or iii) an EV marker. In some embodiments, the EV marker is selected from the group consisting of CD9, CD81 , CD63, TSG101 , CD171 , NrCAM, GLAST, EAAT1 , SYP, ALIX, HSP70, MHC 1 , MHC 2, MOG, SOX10, OLIG1 , IBA1 , TMEM119, TREM2, EAAT2, CD49f, L1 CAM, NCAM, NRXN3, SYN1 , PDGFR beta, CSPG4, CD13, SLC2A1 , and CD31 . In some embodiments, the EV marker is a pan marker of EVs. In some embodiments, the EV marker is exclusive to an EV subpopulation.

Methods of Detecting Extracellular Vesicle-Bound Biomarkers

In some embodiments, the method of detecting EV-associated biomarkers includes an immunoassay. For example, an immunoassay may include an ELISA assay. The ELISA assay may be a direct ELISA, an indirect ELISA, a sandwich ELISA, or a competitive ELISA-based assay.

In some embodiments, a biomarker detection method comprises isolating EVs from a population of cells and first incubating the exogenous EVs with a sample obtained from a patient suspected of disease, followed by detecting the biomarker on the EVs using a binding agent. In some embodiments, the binding agent preferentially (or exclusively binds) to an EV-associated form of the biomarker to a soluble form. In some embodiments, the binding agent binds specifically to the EV-associated form of the biomarker and the soluble form of the biomarker. Obtaining EVs

The method may include an initial step of isolating EVs as described in the above section entitled “Extracellular Vesicles (EVs).” When assaying using the endogenous EVs, the test sample may be tested directly without isolating the endogenous EVs. When assaying using the exogenous EVs, the samples may be incubated with exogenous EVs for a period of time sufficient enough for the EV to act as a sponge. In some embodiments, the exogenous EVs may be isolated from a different patient sample, synthetically synthesized EVs, or EVs from genetically modified cell culture. The EVs may be incubated with the sample for a period of time of from about 5 minutes to about 24 hours. For example, the EVs may be incubated with the sample from about 10 minutes to 30 minutes, from 30 minutes to 60 minutes, from 60 minutes to 90 minutes, from 90 minutes to 120 minutes, from 120 minutes to 150 minutes, from about 150 minutes to about 180 minutes, from about 180 minutes to 210 minutes, from about 210 minutes to about 240 minutes, from about 240 minutes to about 270 minutes, from about 270 minutes to about 300 minutes, from about 300 minutes to about 330 minutes, from about 330 minutes to about 360 minutes, from about 360 minutes to about 420 minutes, from about 420 minutes to about 480 minutes, from about 480 minutes to about 540 minutes, from about 540 minutes to about 600 minutes, from about 600 minutes to about 660 minutes, from about 660 minutes to about 720 minutes, from about 720 minutes to about 780 minutes, from about 780 minutes to about 840 minutes, from about 840 minutes to about 900 minutes, from about 900 minutes to about 960 minutes, from about 960 minutes to about 1020 minutes, from about 1020 minutes to about 1080 minutes, from about 1080 minutes to about 1140 minutes, from about 1140 minutes to about 1200 minutes, from about 1200 minutes to about 1260 minutes, from 1260 minutes to about 1320 minutes, from about 1320 minutes to about 1380 minutes, or from about 1380 minutes to about 1440 minutes.

Assay Configurations

Aspects of the present disclosure include methods of improved sensitivity for detecting biomarkers. In some embodiments, the biomarker is a rare biomarker, e.g., markers present in the test samples are in a low amount such that it is undetectable using conventional methods due to being lower than the limit of detection or quantification of such assays. For example, the methods described herein may act or otherwise perform as a concentration enhancement technique: via the sponge effect of the EVs, the soluble biomarkers that are low in concentration bind to the surface via a multitude of bindings and become more concentrated.

EV-associated biomarkers can be detected by performing binding assays that are well-known in the art. In some embodiments, the methods include an immunoassay. For example, an immunoassay may include an enzyme-linked immunosorbent assay (ELISA). The ELISA assay may be a direct ELISA, an indirect ELISA, a sandwich ELISA, or a competitive ELISA-based assay. An exemplary assay protocol is described in Example 1 . Test samples are incubated with a pair of binding agents comprising a first binding agent and a second binding agent that capture and detect the biomarker of interest. The incubation with the first and second antibodies may be sequential or simultaneous. In some embodiments, the first binding agent may be an antibody, antibody fragment, an aptamer, a peptide, a nucleic acid, a small molecule, or other synthetic agents. The first binding agent may be selected to bind to an EV-associated biomarker or EV marker preferentially. The first antibody can be coated or immobilized on a solid support to capture the EV- associated biomarker or EV marker. In some embodiments, the second binding agent may be coupled to a signal amplification moiety, which can produce a detectable signal, for example, a fluorescent, chemiluminescent, or colorimetric signal. For instance, a fluorophore or an enzyme may generate a detectable signal. In some embodiments, the signal is caused by an enzyme that reacts with a substrate. In some embodiments, the enzyme is a horse radish peroxidase. In some embodiments, the second binding agent may be an antibody conjugated to a fluorophore, which may be readily commercially available. The second binding agent may be selected to bind to an EV-associated biomarker or EV marker preferentially.

After the binding of the second antibody, the signals resulting from the binding can be a fluorescent, chemiluminescent, or colorimetric signal, which can be detected by appropriate detection devices, for example, a spectrometer. The amount of the biomarker of interest, the total EV count, and the total EV subpopulation count can be determined based on the signals.

In some embodiments, the first and second binding agent can selectively bind to a soluble biomarker, an EV-associated biomarker, a soluble and EV-associated biomarker, or an EV marker (FIG. 2). Several configurations can be used to detect the EV-associated biomarkers, and exemplary configurations are shown in Table 1 below. Notably, the binding agents can adopt either configuration when used as a pair. For example, a first binding agent may be UA, and a second binding agent may be EVbA. Alternatively, the first binding agent may be EVbA, and the second binding agent may be UA. The configuration of this pair of antibodies does not affect the measurement (see, e.g., FIGs. 7A-D).

Table 1. Exemplary assay configurations.

Methods Of Detecting Total Extracellular Vesicles

In some embodiments, the method of detecting total EVs in a sample includes a first binding agent and a second binding agent that each selectively bind to an EV marker. In some embodiments, the method of detecting total EVs with an EV subpopulation in a sample includes a first binding agent that can selectively bind to an EV marker and a second binding agent that can selectively bind to an EV-associated form of a source marker characteristic of a cell type of origin for an EV subpopulation. In some embodiments, the method of detecting total EVs with an EV subpopulation in a sample includes a first binding agent that can selectively bind to an EV subpopulation marker and the second binding agent can selectively bind to an EV marker.

In some embodiments, the method of detecting total EVs within a sample includes using light scattering, atomic force microscopy, scanning electron microscopy, flow cytometry, surface plasmon resonance, biolayer interferometry, immunoassays, and lipid/protein staining.

Disease Diagnosis, Monitoring, and Treatment

EVs originating from diverse cell types or tissues showcase distinctive surface markers and membrane compositions, imparting a unique identity to each vesicle type and influencing their interactions with biomarkers. The activation state or overall health of the originating cell is a critical factor shaping the cargo encapsulated within, and at the surface of, EVs. As cells respond to various stimuli or undergo pathological changes, the cargo composition of EVs from different cells becomes a dynamic reflection of the differences in and around specific cells.

The microenvironment enveloping the cell types during the release of EVs plays a pivotal role in defining their cargo. This local environment, which can vary under conditions such as inflammation or cellular stress, contributes to the specific molecular content of the released EVs. Notably, when a particular disease impacts specific cell types, the EVs they release may harbor biomarkers that serve as indicative signatures of the specific pathology. The potential diagnostic and prognostic applications of monitoring disease-induced modifications in EV cargo are considerable. By scrutinizing the molecular contents of EVs from different types of cells, especially those released under pathological conditions, valuable insight into the underlying processes associated with a disease may be obtained.

An aspect of the disclosure described herein includes methods and compositions for diagnosing or monitoring a patient at risk of or currently experiencing a disease and monitoring disease progression. In some embodiments, the disease is a neurological disease, an autoimmune disease, cancer, or inflammation. In some embodiments, the assay may be designed to detect one or more EV-associated biomarkers associated with a neurological disorder, such as CD56, CD171 , NrCAM, EAAT1 , SYP, Tau, pTau, Amyloid beta, alpha-synuclein, phosphorylated alpha-synuclein, TDP43, phosphorylated TDP43, GFAP, NfL, SWOB, NSE, IL-6, IL-1 beta, TNF-alpha, or combinations thereof. In some embodiments, when the assay is designed to detect EV-associated biomarkers associated with autoimmune or inflammation, the biomarkers to be detected may include one or more of CD56, GFAP, IL-6, IL-1 beta, and TNF-alpha. In an alternate embodiment, the assay, designed for cancer diagnosis, detects NSE, TNF-alpha, or both. In another alternate embodiment, the assay, designed for stroke diagnosis, detects NSE, SWOB, or both.

Generating composite EV-associated biomarker values

In some embodiments, a composite value is calculated on the basis of the level of the EV-associated form of a disease biomarker and the total level of EVs within the sample. In some embodiments, the composite value is calculated by normalizing the level of the EV-associated form of a disease biomarker against the level of total EVs in the test sample (or the level of an EV marker that corresponds to the amount of the total EVs in the test sample). In some embodiments, normalization comprises performing a regression on level of the EV-associated form of a disease biomarker against the total EV levels (or the level of an EV marker corresponding to the number of total EVs in the test sample). In some embodiments, a healthy control composite value is obtained on the basis of the level of the EV-associated form of a disease biomarker and the total level of EVs within the healthy control sample. In some embodiments, the healthy control composite value is calculated by normalizing the level of the EV- associated form of a disease biomarker against the level of total EVs in the healthy control sample (or the level of an EV marker that corresponds to the amount of the total EVs in the healthy control sample). In some embodiments, normalization comprises performing a regression on level of the EV-associated form of a disease biomarker against the total EV levels (or the level of an EV marker corresponding to the number of total EVs in the healthy control sample).

In some embodiments, a diseased control composite value is obtained on the basis of the level of the EV-associated form of a disease biomarker and the total level of EVs within the diseased control sample. In some embodiments, the diseased control composite value is calculated by normalizing the level of the EV- associated form of a disease biomarker against the level of total EVs in the diseased control sample (or the level of an EV marker that corresponds to the amount of the total EVs in the healthy control sample). In some embodiments, normalization comprises performing a regression on level of the EV-associated form of a disease biomarker against the total EV levels (or the level of an EV marker corresponding to the number of total EVs in the diseased control sample).

In some embodiments, the level of the EV-associated form of a disease biomarker is measured within an EV subpopulation. In some embodiments, a composite value is calculated on the basis of the level of the EV-associated form of a disease biomarker and the total level of EVs within the EV subpoluation within the sample. In some embodiments, the composite value is calculated by normalizing the level of the EV- associated form of a disease biomarker against the level of total EVs within the EV subpopulation in the test sample (or the level of an EV marker that corresponds to the amount of the total EVs within the EV subpopulation in the test sample). In some embodiments, normalization comprises performing a regression on level of the EV-associated form of a disease biomarker against the total EV levels within the EV subpopulation (or the level of an EV marker corresponding to the number of total EVs within the EV subpopulation in the test sample).

The detected levels of EV-associated biomarker in the disease control group or the composite values thereof may be compared to the detected levels of EV-associated biomarker or the composite values thereof in the healthy control group. If the level of the EV-associated biomarker or the normalized value thereof in the disease sample group is significantly different from the corresponding healthy control group, the data may accurately classify a diseased subject against a healthy subject. A patient may be diagnosed as having a risk of developing or concurrently experiencing the disease on the basis of the classifier and the levels of EV-associated biomarker(s) in the patient’s sample.

Common methods to correct for different amounts of correlation between two biomarkers and create a composite value are: calculation of regression residuals, partial correlation coefficients, covariate adjustment, Z score standardization, adjustment factors (e.g. ratios, log transformations, fixed factors, standardization to reference signals, and other commonly known adjustment factors), or regularization methods. Algorithmic correction and combination of biomarker levels

In some embodiments, an algorithmic correction is performed on the composite values of EV- associated biomarkers in the disease control group and healthy control group for one or more EV-associated biomarkers. In some embodiments, the algorithmic correction includes a classification model and feature selection algorithms. The feature selection algorithms may be utilized to remove non-informative normalized values of EV-associated biomarkers from the classification model to produce a more parsimonious classification model. In some embodiments, a classification model is built on the levels or composite values of EV-associated biomarkers for one or more EV-associated biomarkers to better capture the systemic interactions and/or imbalances in function which can better distinguish the disease control group and healthy control group. In some embodiments, the classification model is non-linear. In some embodiments, the nonlinear classification model is a multiple logistic regression classifier. In some embodiments, the non-linear model includes support vector machines and/or random forests. In some embodiments, the feature selection algorithms include forward selection, recursive feature elimination, and/or penalized regression algorithms. In some embodiments, the classification model yields an accurate disease classifier. In some embodiments, an accurate disease classifier has an AUC of at least 0.6 (e.g., at least 0.65, at least 0.70, at least 0.75, at least 0.80, at least 0.85, at least 0.90, or at least 0.95).

A patient may be diagnosed as having a risk of developing or concurrently experiencing the disease on the basis of the classifier and the level(s) or composite value(s) of EV-associated biomarker(s) in the patient’s sample.

In some embodiments, the prediction score of the biomarker detected in the disease control is significantly different compared to the prediction score of the corresponding healthy control if the p-value is less than 0.05, e.g., less than 0.005 or less than 0.001 .

Subject Diagnosis and Treatment Methods

The methods of the invention can include treating a subject diagnosed using the methods of the invention.

Synucleinopathies

The invention features a method of detecting the presence or level of synuclein aggregation in subjects, including those suffering from a synucleinopathy. Synucleinopathies are characterized by deposition of intracellular protein aggregates that are microscopically visible as Lewy bodies and/or Lewy neurites, where the protein alpha-synuclein is the major component. Synucleinopathies frequently have degeneration of the dopaminergic nigrostriatal system, responsible for the core motor deficits in Parkinsonism (rigidity, bradykinesia, resting tremor). Several non-motor signs and symptoms are thought to precede motor symptoms in Parkinson's disease and other synucleinopathies. Such early signs include, for example, REM sleep behavior disorder and loss of smell and constipation.

Synucleinopathies include Parkinson's disease (PD) (including idiopathic and inherited forms of Parkinson's disease as well as prodromal PD) and Diffuse Lewy Body (DLB) disease (also known as Dementia with Lewy Bodies (DLB), multiple system atrophy (MSA; e.g., Olivopontocerebellar Atrophy, Striatonigral Degeneration, and Shy-Drager Syndrome)), Lewy body variant of Alzheimer's disease, combined Alzheimer's and Parkinson disease, and pure autonomic failure. In some embodiments, the synucleinopathy is not Alzheimer's disease.

In some embodiments, the synucleinopathy is PD. In some embodiments, the PD is a subtype of the disease thereof (e.g., prodromal PD). In some embodiments, the synucleinopathy is DLB. In some embodiments, the synucleinopathy is MSA.

The present invention further provides methods for treating a patient suffering from a synucleinopathy (e.g., DLB, MSA, PD or a subtype of the disease thereof e.g., prodromal PD). In some instances, the methods of the invention includes administering to the patient an effective amount of a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and/or a neuroprotective agent. Any of the cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), anti-tremor agents, and neuroprotective agents described herein or otherwise known in the art may be used in the methods. In some instances, the methods involve (i) diagnosing the subject as having a synucleinopathy by analyzing a plasma sample obtained from a subject as described herein and, following step (i), administering a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and/or a neuroprotective agent to the subject.

The invention features a method of treating a subject suffering from a synucleinopathy (e.g., DLB, MSA, PD or a subtype of the disease thereof e.g., prodromal PD), the method including monitoring pathogenesis and/or response to therapy.

In any of the preceding methods, the cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), anti-tremor agent, and/or neuroprotective agent may be any cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), anti-tremor agent, and/or neuroprotective agent known in the art or described herein.

The compositions used in the methods described herein (e.g., a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), an anti-tremor agent, and/or a neuroprotective agent) can be administered by any suitable method, including, for example, intravenously, intramuscularly, subcutaneously, orally, by injection, by implantation, or by infusion. The compositions utilized in the methods described herein can also be administered systemically or locally. The method of administration can vary depending on various factors (e.g., the compound or composition being administered and the severity of the condition, disease, or disorder being treated). Dosing can be by any suitable route, e.g., by injections, such as intravenous or subcutaneous injections, depending in part on whether the administration is brief or chronic. Various dosing schedules including but not limited to single or multiple administrations over various time-points, bolus administration, and pulse infusion are contemplated herein.

Cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), antitremor agents, and neuroprotective agents described herein may be formulated, dosed, and administered in a fashion consistent with good medical practice. Factors for consideration in this context include the particular disease subtype being treated (e.g., PD and prodromal PD, MSA, and DLB), the clinical condition of the individual patient, the cause of the disease, the site of delivery of the agent, the method of administration, the scheduling of administration, and other factors known to medical practitioners. Exemplary cognition-enhancing agents useful in the methods of the invention include, but are not limited to, donepezil, rivastigmine tartrate, galantamine HBr, memantine, and modafinil.

Exemplary antidepressant agents useful in the methods of the invention include sertraline, fluoxetine, citalopram, escitalopram, paroxetine, and fluvoxamine.

Exemplary anxiolytic agents useful in the methods of the invention include, but are not limited to, alprazolam, chlordiazepoxide, clobazepam, clonazepam, clorazepate, diazepam, estazolam, and flurazepam.

Exemplary antipsychotic agents useful in the methods of the invention include, but are not limited to, aripiprazole, asenapine, cariprazine, clozapine, lurasidone, olanzapine, quetiapine, and risperidone.

Exemplary sedatives useful in the methods of the invention include, but are not limited to, alprazolam, chloral hydrate, chlordiazepoxide, clorazepate, clonazepam, diazepam, and estazolam.

Exemplary dopamine promoters (agonists) useful in the methods of the invention include, but are not limited to, selegiline, pramipexole and levodopa (L-DOPA).

Exemplary anti-tremor agents useful in the methods of the invention include, but are not limited to, propranolol, primidone, gabapentin, and topiramate.

Exemplary neuroprotective agents useful in the methods of the invention include, but are not limited to, gangliosides, topiramate, riluzole, methylprednisolone, rivstigmine, selegiline, cilostazol, rasagiline, tenocyclidine, 7-nitroindazole, N-(3-propylcarbamoyloxirane-2-carbonyl)-isoleucyl-proline, huperzine A, SGS- 742, D-JNKI-1 , nalmefene, ziconotide, dexanabinol, remacemide, clomethiazole, propentofylline, Z-Val-Ala- Asp fluoromethyl ketone, piracetam, epigallocatechin gallate, vinpocetine, tempol, butylphthalide, eliprodil, tirilazad, nefiracetam, gacyclidine, nizofenone, meclofenoxate, linopiridine, fosfructose, methylprednisolone hemisuccinate, dextrorphan, ebselen, almitrine, brimapitide, edaravone, edaravone, minocycline, epoetin-p, trafermin, filgrastim, eicosapentaenoic acid, and pioglitazone.

Vascular Damage

The invention features a method of detecting the presence or level of vascular damage in subjects, including those suffering from hemorrhagic or occlusive vascular damage, and those suffering from brain vascular damage, such as a stroke. Where brain vascular damage is indicated, the method can further include the step of performing a brain scan to distinguish between ischemic and hemorrhagic stroke.

Subjects suffering from ischemic stroke can further receive reperfusion therapy (e.g., by administration of tPA or other thrombolytic agents or by mechanical devices that increase blood flow to the affected area) and/or agents that treat the damaging effects of ischemia on, e.g., the central nervous system or assist in reperfusion of the ischemic tissue. Optionally, the device is a coil, stent, balloon (eg, intra-aortic balloon, pump), or catheter. Optionally, the agent is a thrombolytic agent (e.g., streptokinase, acylated plasminogen-streptokinase activator complex (APSAC), urokinase, single- chain urokinase-plasminogen activator (scu-PA), anti-inflammatory agents, vasodilator, hypertensive drug, an anticoagulant (e.g., warfarin or heparin); antiplatelet drug (e.g., aspirin); a glycoprotein llb/llla inhibitor; a glycosaminoglycan; coumarin; GCSF; melatonin; an apoptosis inhibitor (e.g., caspase inhibitor), an anti-oxidant (e.g., NXY-059); and a neuroprotectant (e.g., an NMDA receptor antagonists or a cannabinoid antagonist). Exemplary neuroprotective agents useful in the methods of the invention include, but are not limited to, gangliosides, topiramate, riluzole, methylprednisolone, rivstigmine, selegiline, cilostazol, rasagiline, tenocyclidine, 7- nitroindazole, N-(3-propylcarbamoyloxirane-2-carbonyl)-isoleucyl-proline, huperzine A, SGS-742, D-JNKI-1 , nalmefene, ziconotide, dexanabinol, remacemide, clomethiazole, propentofylline, Z-Val-Ala-Asp fluoromethyl ketone, piracetam, epigallocatechin gallate, vinpocetine, tempol, butylphthalide, eliprodil, tirilazad, nefiracetam, gacyclidine, nizofenone, meclofenoxate, linopiridine, fosfructose, methylprednisolone hemisuccinate, dextrorphan, ebselen, almitrine, brimapitide, edaravone, edaravone, minocycline, epoetin-p, trafermin, filgrastim, eicosapentaenoic acid, and pioglitazone.

Subjects suffering from hemorrhagic stroke can further receive acute blood pressure management (e.g., by administering beta-blockers, ACE inhibitors, calcium channel blockers, and/or hydralazine), coagulopathy reversal (e.g., vitamin K combined with prothrombin complex concentrates (PCCs); and/or thrombin inhibitors and factor Xa inhibitors (FXa-ls), such as idarucizumab and andexanet alfa), and surgical hematomaevacuation. In one particular embodiment, the subject is treated with an anticoagulant reversal agent selected from the group consisting of protamine, protamine sulfate, vitamin K, prothrombin complex concentrate (PCC), idarucizumab, Andexanet Alfa, and combinations thereof.

Tau pathologies

The invention features a method of detecting the presence or level of tau aggregation in subjects, including those suffering from canonical and non-canonical tauopathies. Tauopathies are neurodegenerative disorders characterized by the deposition of abnormal tau protein in the brain. The spectrum of tau pathologies expands beyond the traditionally discussed disease forms like Pick disease, progressive supranuclear palsy, corticobasal degeneration, and argyrophilic grain disease. Emerging entities and pathologies include globular glial tauopathies, primary age-related tauopathy, which includes neurofibrillary tangle dementia, chronic traumatic encephalopathy (CTE), and aging-related tau astrogliopathy. Clinical symptoms include frontotemporal dementia, corticobasal syndrome, Richardson syndrome, parkinsonism, pure akinesia with gait freezing and, rarely, motor neuron symptoms or cerebellar ataxia.

The present invention further provides methods for treating a patient suffering from a tauopathy. In some instances, the methods of the invention includes administering to the patient an effective amount of a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), and/or a neuroprotective agent. Any of the cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), and neuroprotective agents described herein or otherwise known in the art may be used in the methods. In some instances, the methods involve (i) diagnosing the subject as having a tauopathy by analyzing a plasma sample obtained from a subject as described herein and, following step (i), administering a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), and/or a neuroprotective agent to the subject.

The invention features a method of treating a subject suffering from a tauopathy, the method including monitoring pathogenesis and/or response to therapy.

In any of the preceding methods, the cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), and/or neuroprotective agent may be any cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), and/or neuroprotective agent known in the art or described herein. The compositions used in the methods described herein (e.g., a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), and/or a neuroprotective agent) can be administered by any suitable method, including, for example, intravenously, intramuscularly, subcutaneously, orally, by injection, by implantation, or by infusion. The compositions utilized in the methods described herein can also be administered systemically or locally. The method of administration can vary depending on various factors (e.g., the compound or composition being administered and the severity of the condition, disease, or disorder being treated). Dosing can be by any suitable route, e.g., by injections, such as intravenous or subcutaneous injections, depending in part on whether the administration is brief or chronic. Various dosing schedules including but not limited to single or multiple administrations over various timepoints, bolus administration, and pulse infusion are contemplated herein.

Cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), and neuroprotective agents described herein may be formulated, dosed, and administered in a fashion consistent with good medical practice. Factors for consideration in this context include the particular disease subtype being treated, the clinical condition of the individual patient, the cause of the disease, the site of delivery of the agent, the method of administration, the scheduling of administration, and other factors known to medical practitioners.

Exemplary cognition-enhancing agents useful in the methods of the invention include, but are not limited to, donepezil, rivastigmine tartrate, galantamine HBr, memantine, and modafinil.

TDP43 diseases

The invention features a method of detecting the presence or level of TDP43 aggregation in subjects, including those suffering from standard and non-standard TDP43 diseases. Inclusions of pathogenic deposits containing TAR DNA binding protein 43 (TDP43) are evident in the brain and spinal cord of patients that present across a spectrum of neurodegenerative diseases. For instance, the majority of patients with sporadic amyotrophic lateral sclerosis (up to 97%) and a substantial proportion of patients with frontotemporal lobar degeneration (-45%) exhibit TDP-43 positive neuronal inclusions, suggesting a role for this protein in disease pathogenesis. In addition, TDP-43 inclusions are evident in familial ALS phenotypes linked to multiple gene mutations including the TDP-43 gene coding (TARDBP) and unrelated genes (eg, C9orf72).

The present invention further provides methods for treating a patient suffering from a TDP43 disease. In some instances, the methods of the invention includes administering to the patient an effective amount of a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), TDP43 therapy, and/or a neuroprotective agent. Any of the cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), and neuroprotective agents described herein or otherwise known in the art may be used in the methods. In some instances, the methods involve (i) diagnosing the subject as having a TDP43 disease by analyzing a plasma sample obtained from a subject as described herein and, following step (i), administering a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), TDP43 therapy, and/or a neuroprotective agent to the subject.

The invention features a method of treating a subject suffering from a TDP43 disease, the method including monitoring pathogenesis and/or response to therapy.

In any of the preceding methods, the cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), and/or neuroprotective agent may be any cognition-enhancing agent, antidepressant agent, dopamine promoter (e.g., agonist), TDP43 therapy, and/or neuroprotective agent known in the art or described herein.

The compositions used in the methods described herein (e.g., a cognition-enhancing agent, an antidepressant agent, a dopamine promoter (e.g., agonist), TDP43 therapy, and/or a neuroprotective agent) can be administered by any suitable method, including, for example, intravenously, intramuscularly, subcutaneously, orally, by injection, by implantation, or by infusion. The compositions utilized in the methods described herein can also be administered systemically or locally. The method of administration can vary depending on various factors (e.g., the compound or composition being administered and the severity of the condition, disease, or disorder being treated). Dosing can be by any suitable route, e.g., by injections, such as intravenous or subcutaneous injections, depending in part on whether the administration is brief or chronic. Various dosing schedules including but not limited to single or multiple administrations over various time-points, bolus administration, and pulse infusion are contemplated herein.

Cognition-enhancing agents, antidepressant agents, dopamine promoters (e.g., agonists), TDP43 therapeutics, and neuroprotective agents described herein may be formulated, dosed, and administered in a fashion consistent with good medical practice. Factors for consideration in this context include the particular disease subtype being treated, the clinical condition of the individual patient, the cause of the disease, the site of delivery of the agent, the method of administration, the scheduling of administration, and other factors known to medical practitioners.

Exemplary TDP43 therapeutics useful in the methods of the invention include, but are not limited to, riluzole, edaravone, AMX0035, sodium phenylbutyrate, taurursodiol, tofersen, quinidine sulfate, dextromethorphan hydrobromide, and Neudexta (e.g., a combination of quinidine sulfate and dextromethorphan hydrobromide). Amyloid beta diseases

The invention features a method of detecting the presence or level of amyloid beta aggregation in subjects, including those suffering from amyloid beta diseases. Amyloid beta primarily refer to a group of neurodegenerative disorders characterized by the accumulation of amyloid beta peptides in the brain. The most prominent of these diseases is Alzheimer's disease, but there are other conditions associated with amyloid pathology such as Cerebral Amyloid Angiopathy, and Frontotemporal dementia with amyloid pathology.

The present invention further provides methods for treating a patient suffering from an amyloid beta disease. In some instances, the methods of the invention includes administering to the patient an effective amount of a cognition-enhancing agent or an amyloid beta aggregate targeting agent. Any of the cognitionenhancing agents described herein or otherwise known in the art may be used in the methods. In some instances, the methods involve (i) diagnosing the subject as having an amyloid beta disease by analyzing a plasma sample obtained from a subject as described herein and, following step (i), administering a cognitionenhancing agent to the subject. The invention features a method of treating a subject suffering from an amyloid beta disease, the method including monitoring pathogenesis and/or response to therapy. Exemplary cognition-enhancing agents useful in the methods of the invention include, but are not limited to, donepezil, rivastigmine tartrate, galantamine HBr, memantine, and modafinil. Exemplary cognition-enhancing agents useful in the methods of the invention include, but are not limited to, Lecanemab, Aducanumab, Donanemab, and Gantenerumab.

Exogenous EV Methods

In yet another aspect, diagnosing or monitoring a disease includes using exogenous EVs to detect a biomarker associated with a disease. The method comprises isolating exogenous EVs from a source and incubating the isolated EVs with a test sample obtained from a patient, for example, a patient diagnosed with a particular disease for which the biomarker is an indicator.

In some embodiments, the methods described herein may include assays designed to detect a diseased state using exogenous EVs. In some embodiments, the assays may be designed to utilize exogenous EVs isolated from different subpopulations of cells, thereby providing different subpopulations of EVs. For example, EVs may be isolated from brain cells, endothelial cells, epithelial cells, glial cells, astrocyte cells, pericytes, vascular endothelial cells, or neuronal cells and thereby incubated with the blood sample to obtain information on the different concentrations of biomarker present depending upon the EV subpopulation exogenously loaded.

In some embodiments, the exogenous EVs may be isolated from a patient already diagnosed with or experiencing a disease. The exogenous EVs may provide insight into disease progression in a patient based on the expression of biomarkers measured in the sample. In some embodiments, a patient sample may be assayed at a first time point wherein no treatment for a condition has been performed and later assayed at a second time point after a treatment has been completed. In some embodiments, the patient may undergo repeat treatment cycles while monitoring the EVs in a sample to evaluate the progression of the disease during treatment. The method comprises isolating exogenous EVs from a source. For example, the exogenous EVs may be isolated from a different patient sample, synthetically synthesized EVs, or EVs from genetically modified cell culture (e.g., neuronal cells, astrocyte cells, or epithelial cells). In some embodiments, the binding of biomarkers to the EVs may depend on the source of exogenous EVs. For example, EVs isolated from neuronal cells may bind biomarkers that are different from biomarkers that may bind to EVS isolated from astrocyte cells in a patient sample. The EVs may be incubated with the sample for a period of time of from about 5 minutes to about 24 hours. For example, the EVs may be incubated with the sample from about 10 minutes to 30 minutes, from 30 minutes to 60 minutes, from 60 minutes to 90 minutes, from 90 minutes to 120 minutes, from 120 minutes to 150 minutes, from about 150 minutes to about 180 minutes, of from about 180 minutes to 210 minutes, of from about 210 minutes to about 240 minutes, of from about 240 minutes to about 270 minutes, of from about 270 minutes to about 300 minutes, of from about 300 minutes to about 330 minutes, of from about 330 minutes to about 360 minutes, of from about 360 minutes to about 420 minutes, of from about 420 minutes to about 480 minutes, of from about 480 minutes to about 540 minutes, of from about 540 minutes to about 600 minutes, of from about 600 minutes to about 660 minutes, of from about 660 minutes to about 720 minutes, of from about 720 minutes to about 780 minutes, of from about 780 minutes to about 840 minutes, of from about 840 minutes to about 900 minutes, of from about 900 minutes to about 960 minutes, of from about 960 minutes to about 1020 minutes of from about 1020 minutes to about 1080 minutes, of from about 1080 minutes to about 1140 minutes, of from about 1140 minutes to about 1200 minutes, of from about 1200 minutes to about 1260 minutes of from 1260 minutes to about 1320 minutes, of from about 1320 minutes to about 1380 minutes, or of from about 1380 minutes to about 1440 minutes.

Following incubation, the assay using exogenous EVs may include coating the first binding agent (for example, a first antibody) to a solid support and incubating a test sample from a patient to be diagnosed with a first binding agent on the solid support. The first binding agent may be a UA, an EVbA, or an EVA, as shown in Table 1. The patient sample containing EVs may then be added to the solid support platform and allowed to bind for a sufficient time to allow the EVs to bind to the first binding agent. For example, the patient sample may be previously frozen or collected from a patient and directly incubated onto the solid support. In some embodiments, the patient samples were processed via the methods described below before being used in the assay. In some embodiments, the incubation time may be from 10 minutes to about 6 hours. For example, the patient sample may be incubated for about 10 minutes to 30 minutes, from 30 minutes to 60 minutes, from 60 minutes to 90 minutes, from 90 minutes to 120 minutes, from 120 minutes to 150 minutes, from about 150 minutes to about 180 minutes, of from about 180 minutes to 210 minutes, of from about 210 minutes to about 240 minutes, of from about 240 minutes to about 270 minutes, of from about 270 minutes to about 300 minutes, of from about 300 minutes to about 330 minutes, of from about 330 minutes to about 360 minutes.

After incubation with the blood sample, the assay may be washed with a buffer solution to allow unbound EVs and protein to wash away from the surface. A second binding agent, diluted in a buffer, for example, a buffer containing BSA and/or PBST, is then added to the assay mixture. In some embodiments, the second binding agent (for example, an antibody) may be a UA, an EVbA, or an EVA, as shown in Table 1. In some embodiments, the incubation time for the second antibody may be from 10 minutes to about 2 hours. For example, the patient sample may be incubated for about 10 minutes, for about 20 minutes, for about 30 minutes, for about 40 minutes, for about 50 minutes, for about 60 minutes, for about 70 minutes, for about 80 minutes, for about 90 minutes, for approximately 100 minutes, for approximately 110 minutes or about 120 minutes. The secondary antibody may be bound to either horse radish peroxidase or horse radish peroxidase conjugated with streptavidin.

The assay may further be incubated with a chemiluminescent substrate, and an intensity measurement was performed to quantify the biomarker signal. In parallel to the steps described above, a control sample may be treated with the same steps in connection with the described assay. For example, a sample from the same tissue origin in a healthy individual may be obtained as the control sample, and the sample may be processed via the steps described herein. The signal intensity of the healthy control is compared to the signal intensity of the patient sample to determine the patient’s disease status.

In some embodiments, the assay for detecting a disease may be formatted by pairing UA (first antibody) with EV-associated antibody (EVbA, second antibody), UA (first antibody) with EVA (second antibody), EVbA (first antibody) with UA (second antibody), EVbA (first antibody) with EVbA (second antibody), EVbA (first antibody) with EVA(second antibody), EVA (first antibody) with UA (second antibody), or EVA (first antibody) with EVbA (second antibody). Various configurations of the assays can be used to detect the EV-associated biomarkers, are shown in FIG. 2 and Table 1.

EXAMPLES

The following examples are put forth so as to provide those of ordinary skill in the art with a description of how the compositions and methods described herein may be used, made, and evaluated, and are intended to be purely exemplary of the disclosure and are not intended to limit the scope of what the inventors regard as their disclosure.

Example 1. Methods

Plasma/Serum Separation

Venous blood was drawn and collected in either EDTA or citrate or heparin-treated tubes or serum separator tubes for plasma collection. To isolate plasma, whole blood collected in the appropriate tube was centrifuged for 10 minutes at 400 x g to 1 ,000 x g. Plasma was transferred to a separate low-binding microcentrifuge tube. The plasma was centrifuged for 10 minutes at 1 ,100 x g to produce platelet-depleted plasma. Serum samples were prepared by allowing the whole blood to clot in the serum separator tube for 30 minutes at room temperature. The clot was removed by centrifuging at 1 ,000 x g for 10 minutes, and serum was transferred to a separate low-binding microcentrifuge tube. Aliquoted plasma or serum samples were stored at -80°C before measurements.

Cell Culture

Human cell lines SH-SY5Y (neuron), GLI36 (glioblastoma), and A431 (epithelial carcinoma) were used. GLI36 and A431 were grown in Dulbecco’s Modified Essential Medium (DMEM, Gibco). SH-SY5Y was cultured in Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12 Medium (DMEM/F12 Gibco). All media were supplemented with 10% FBS and penicillin-streptomycin. Recombinant Extracellular Vesicle (EV) expression

Cells at passages 1-15 were cultured in a vesicle-depleted medium (with 5% depleted FBS) for 48 hours before vesicle collection. All media containing EVs were filtered through a 0.2pm membrane filter (regenerated cellulose, Millipore), or plasma, serum, ascites, or urine containing EVs were isolated by differential centrifugation (first at 10,000 x g and subsequently at 100,000 x g), and used for EV analysis. The nanoparticle tracking analysis (NTA) system (NS300, Nanosight) for independent quantifications of EV concentration was utilized. To achieve optimal counting, EV concentrations were adjusted to obtain ~50 vesicles in the field of view. All NTA measurements were done with identical system settings for consistency.

Recombinant protein expression

For the expression of the target protein, a recombinant plasmid was constructed containing the gene of interest along with any purification tags (e.g., HIS, GST, FLAG, and the like). The final plasmid sequence was confirmed with sequencing of the full plasmid. The plasmid was transfected into bacteria (E. Coli), yeast (S. Cerevisiae), or mammalian (CHO) cells using either electroporation or lipofection. Protein expression was induced under controlled conditions using an induction chemical (e.g., IPTG, tet, and the like) or constitutive expression. Cells were harvested at the optimal time post-induction. Harvested cells underwent lysis buffer (Thermo Fisher Scientific) treatment to release the expressed proteins from the cells. The target protein was purified through affinity chromatography using commercial kits (Thermo Fisher Scientific). The purity and concentration of the purified protein were assessed using SDS-PAGE and Mass Spectrometry.

Enzyme-Linked ImmunoSorbent Assay (ELISA)

Between each step of the ELISA, the plates were washed with PBST (PBS with 0.05% Tween® 20). Capture antibodies (0.1 - 5 pg/ml) were adsorbed onto ELISA plates (Thermo Fisher Scientific) and then blocked with BSA (0.01% - 5%). Samples were diluted in appropriate diluent (e.g., PBS, 0.1% BSA, 0.2% - 2% Triton™ X-100, and the like), added to each well, and incubated for 1-6 hours at room temperature. Detection antibodies (0.1 - 5 pg/ml) in 0.01% - 1% BSA, PBST were added and incubated for 1 hour at room temperature. Following incubation with either horseradish peroxidase-conjugated secondary antibody or horseradish peroxidase-conjugated streptavidin, a chemiluminescent substrate was added, and intensity was used to quantify biomarker signal (Tecan).

Scanning Electron Microscopy (SEM)

Samples were diluted in PBS and incubated on an antibody-functionalized, pre-blocked surface for 1 hour before being rinsed with PBST (PBS + 0.05% Tween® 20). After dehydration in a series of increasing ethanol concentrations, samples were transferred for critical point drying (Leica) and subsequently sputter- coated with gold (Leica) before imaging with a scanning electron microscope (JEOL 6701 ).

Assay configurations

To characterize the state of the biomarker recognized by the different antibodies, recombinant proteins were used as standards for the soluble biomarker forms, exogenous EVs as standards for total EV, and a pooled plasma/serum sample as a standard for EV-associated forms of biomarkers. Each standard was tested in different dilutions to ensure the specificity of the antibody. EV assays were designed using an EV-specific binding agent (recognizing EV markers) (EVA) paired with EVA.

Universal biomarker assays (soluble and EV-associated forms of the biomarker) were designed using a universal binding agent (UA) paired with UA.

Soluble form-specific biomarker assays were designed using a soluble specific binding agent (SSA) paired with SSA, SSA paired with UA, or UA paired with SSA.

EV-associated specific biomarker assays were designed by pairing UA with EV-associated binding agent (EVbA), UA with EVA, EVbA with UA, EVbA with EVbA, EVbA with EVA, EVA with UA, or EVA with EVbA.

For different biomarkers, depending on the availability of the antibody, various configurations were used to get the assays specific to various biomarker forms. Exemplary assay configurations for measures of soluble biomarker, soluble and EV-associated biomarker, EVs, and EV-associated biomarkers only (no soluble form) are shown in FIG. 2. First generation assays utilized an antibody pair including i) an EVbA or UA and ii) and EVbA to measure EV-associated biomarkers. Second generation assays utilized an antibody pair including i) an EVbA or UA and ii) an EVA to measure EV-associated biomarkers, improving the certainty that the levels measured with the second generation assay reflect EV-assocaiated disease markers (as compared to the soluble form of the disease marker), while also providing the opportunity to measure EV-associated disease markers from a particular cellular origin (e.g., using an EVA specific to neuronal EVs). The level of total EVs for both the first generation and second generation assays utilized two EVAs. The first generation assay utlized the same two EVAs, whereas, the second generation, optionally, utilized two different EVAs to measure the level of total EVs in the sample or the level of total EVs within a subpopulation having a specific cellular origin.

Sample filtration

Samples were passed through a 0.8 pm or 0.2 pm membrane filter or a 10OkDa spin column filter (regenerated cellulose, Millipore). The filtrate and retentate were recovered and used to quantify biomarker signals using the above methods.

Clinical sample collection and measurements

The study was approved by the Parkway Independent Ethics Committee (PIEC). Subjects were recruited from multiple independent cohorts and were either clinically diagnosed by a trained neurologist or diagnosed using PET brain imaging results read by a trained radiologist. For plasma collection, venous blood was collected from subjects in EDTA tubes and processed within 8 hours of collection using the methods described above. Soluble and/or EV-associated forms of each biomarker, and EV quantity were measured on native plasma/serum samples without vesicle isolation.

Regression and statistical analysis

Descriptive statistics, including means, standard deviations, and frequencies, were calculated for each variable. The error bars within the Figures represent technical variability (i.e. , varaibility among replicates in the same experimental run). Univariate analyses were conducted to identify significant associations between individual predictors and the outcome. Further method detailing generating composite values and utilizing algorithmic correction are later described in Example 10.

Example 2. Amplification of Biomarker Signals for Rare Biomarkers and Biomarkers Relevant to Specific Diseases by Measuring the EV-associated Form of the Biomarker

Using binding agents identified as described above, assays were designed to detect EV-associated biomarker epitopes. This complex interplay leads to accumulating a broad spectrum of biomarkers both within and on EVs, substantially enhancing the signal of rare circulating biomarkers by up to 10,000 times. In the case of commonly circulating biomarkers, the EV-associated form might be less prevalent than the soluble form. However, the interaction between EVs and biomarkers more accurately reflects the microenvironment of the parent cell. The enhanced fidelity to the parent cell’s surroundings can allow for greater discrimination between individuals with diseases and those who are healthy (FIGs. 3B and 3C).

Specifically, three assays were utilized to test their ability to bind soluble and EV-associated biomarkers. Samples were first measured for endogenous EVs only, and subsequently, samples incubated with exogenous EVs were also measured. The exogenous EV spiked samples were incubated for 60 minutes before immunoassay analysis. Specifically, three assays to detect a disease in rare biomarkers (FIG. 3B) and relevant forms of biomarkers (FIG. 3C), were performed in parallel. The first assay was developed to detect only the soluble form of the biomarker (using soluble specific antibodies SSA) while the second utilized antibodies that were directed towards the EV-associated biomarker (EVbA), and the third platform utilized universal antibodies (UA) that target both soluble biomarker and EV-associated biomarker. The signals were normalized against a reference standard protein or EV signal to yield the same scale for the protein and EV measurements. For rare biomarkers, the normalized signals detected were predominately EV-associated compared to their soluble form. Using the assay designed to bind soluble and EV-associated biomarkers, the disease controls had higher normalized signals as compared to the healthy controls (FIG. 3B)

A second set of assays were performed to evaluate their ability to detect different biomarker properties (e.g., whether or not the biomarker associates with EVs). Samples evaluated in assays in this format showed a higher proportion of soluble biomarkers than the respective EV-associated biomarkers. When comparing the healthy vs disease samples, there was no difference between the sample cohorts when measuring only the soluble form. In contrast, the assays designed for either the EV-associated biomarker only or for both the EV-associated biomarker and soluble forms demonstrated higher normalized signals in the disease samples compared to the healthy (FIG. 3C).

Example 3. Identifying Enhanced Normalized Signals from EV-associated Forms vs Soluble Forms of Biomarkers

To examine the EV-associated biomarker vs soluble biomarker assay formats, a single patient sample was tested for CD9, CD63, CD81 , CD56, CD171 , NrCAM, EAAT1 , SYP, Tau, phosphorylated Tau, Amyloid beta, Amyloid Beta 42, alpha-synuclein, phosphorylated alpha-synuclein, TDP43, phosphorylated TDP43, GFAP, NfL, S100B, NSE, IL-6, IL-1 beta, and TNF-alpha. CD9, CD63, CD81 , CD56, and CD171 have large quantities of the soluble form compared to the other biomarkers. The levels measured for each soluble and EV-associated form of the biomarker were normalized against a reference standard signal. Interestingly, the normalized signal generated by the EV-associated biomarker was higher than all soluble forms of the biomarkers tested other than CD9 and CD171 . For example, in the blood sample, CD63, CD81 , CD56, alpha-synuclein, IL-6, IL-1 beta, and TNF-alpha had larger normalized signals as compared to the soluble form (FIG. 4).

EV interactions with biomarkers appear to be universal, encompassing a diverse array of biomarkers and types of EVs. Identified above, both soluble and EV-associated forms of biomarker cargo that are known to be associated with specific processes of neurology (alpha-synuclein, phosphorylated alpha-synuclein, tau, phosphorylated tau, amyloid beta isoforms, TDP43, phosphorylated TDP43, GFAP, neurofilament light, SWOB, NSE), protein aggregation (alpha-synuclein, phosphorylated alpha-synuclein, tau, phosphorylated tau, amyloid beta isoforms, TDP43, phosphorylated TDP43), inflammatory (GFAP, IL-6, IL-1 beta, TNF- alpha), and oncology (NSE, TNF-alpha) and many known EV markers (CD9, CD63, CD81 , CD56, CD171 , NrCAM, EAAT1 , SYP), were identified.

The enrichment of rare soluble biomarkers was observed across a variety of biomarkers, highlighting the utility of EVs in concentrating and detecting these rare biomarkers (phosphorylated alpha-synuclein, Tau, phosphorylated tau, amyloid beta isoforms, TDP43, phosphorylated TDP43, GFAP, neurofilament light, SWOB, NSE, NrCAM, EAAT1 , SYP). However, the findings also revealed an intriguing scenario with several biomarkers, where both the soluble and EV-associated forms are highly prevalent, demonstrating a distinct molecular profile from its soluble counterpart (CD9, CD63, CD81 , CD56, CD171 , alpha-synuclein, IL-6, IL- 1 beta, TNF-alpha) (FIG. 4).

To further validate the assay platforms, 30 patient plasma samples were used in the soluble CD171 - specific assay, the EV-associated CD171 -specific assay, and the EV assay platforms (FIG. 5A). The CD171 assay representing high amounts of the soluble form in blood was compared to the Tau-specific assay representing low amounts of soluble form in blood (FIG. 5B). The results demonstrate that no two patients are alike when determining the concentration of the biomarker. For example, the soluble form of CD171 was inconsistent across all 30 patient samples. Interestingly, the EV-associated CD171 appeared to correlate to the sample's EV quantity. Thus, the concentration of the EV-associated biomarker may be standardized to EV concentration. The Tau specific assay also demonstrated similar results (e.g., the EV-associated form of the biomarker scales proportional to the patient's endogenous EV production levels, but the soluble form does not). It is hypothesized that the EVs act as a scaffold or sponge for the biomarker in the parent cell microenvironment. Thus, the more endogenous EVs that cells produce, the more the biomarkers associate with the generated endogenous EVs, but the amount of biomarker per endogenous EV can be informative of the parent cell microenvironment.

Example 4. Assays Demonstrate High Specificity Toward Target Despite Binding Agent Pair Configuration

A set of assays were designed to evaluate the specificity of the assays toward their target recombinant protein (FIG. 6). The target recombinant protein assays utilized UA affinity agents. Soluble reference standards were measured with assays utilizing SSA affinity agents. Brain pathologic, brain function, and EV marker assays were designed with antibody pairs used to measure the following recombinant proteins: Amyloid beta-40 (Ap40), Amyloid beta-42 (Ap42), Tau-441 (2N4R), GSK3-beta phosphorylated Tau (pTau), alpha-synuclein (aSyn), PLK2 phosphorylated alpha-synuclein (paSyn), GFAP, NfL, NSE, S100B, CD9, CD63, CD81 , CD56, CD171 , and NrCAM. High normalized signals were obtained for each assay’s target recombinant proteins, while little to no signal was obtained for other recombinant proteins. Accordingly, the assays demonstrated high specificity for their target with no cross-reactivity against any other recombinant protein target. Total amyloid beta, Tau, and alpha synuclein assays are able to bind to all isoforms of the protein, while protein form specific assays are specific only to its reported isoform/post-translational modification.

In a second experiment, immunoassays were designed to assess the selectivity toward an EV- associated biomarkers using different configurations of binding agent pairs. The first configuration utilized an EVbA affinity agent as the capture agent, while an EVA affinity agent was used as the detection agent. The second configuration utilized the same EVbA affinity agent as the detection agent, while the same EVA affinity agent was used as the capture agent. An additional assay for each configuration was designed to demonstrate the impact of the EV marker, CD56 or CD81 , on target specificity. Two sets of the four assays were designed for measuring the absolute concentration of EV-associated Tau and EV-associated alpha synuclein. The plasma of 11 subjects was tested across the four assays for the two types of biomarkers (FIGs. 7A-D). Among the eight assays, there was a strong correlation of signal between the two configurations, demonstrating that the binding agent pairs, despite their configuration, are each specific to their respective target.

Example 5. Characterization of EV-associated Biomarkers

The size of soluble versus EV-associated biomarkers was evaluated to corroborate the hypothesis that the elevated concentrations of the EV-associated biomarkers were related to the phenomenon of the sponge effect (FIG. 8). Initially, samples were systematically passed through filter of various pore sizes (800 nm, 220 nm, and 100 kDa (about 3 nm)). The biomarker signals were then measured in both the filtrates and retentates. 1 mL of human plasma was initially spiked with soluble Tau recombinant protein before being filtered, and the subsequent retentate and filtrate were assayed for both soluble and EV-associated Tau protein. Passing the human plasma through an 800 nm filter, the retentate had little to no expression of EVs, EV-associated Tau, or soluble Tau, while the filtrate had high expression of EVs, EV-associated Tau, and soluble Tau. The absolute concentrations measured for each soluble and EV-associated form of Tau were normalized against a reference standard. Similar results were found when passing the sample through a 220 nm filter. As expected, after passing the sample through the 100 kDa filter, the retentate had a high concentration of EVs and EV-associated Tau while little to no concentration of soluble Tau. The soluble Tau was only detected in the filtrate, while both EV marker (CD9) and EV-associated Tau were undetected.

In the second experiment, antibodies specific to either EV-associated or soluble forms were coated onto a flat surface to capture molecules from a human plasma sample (FIG. 9). The captured molecules were then visualized using Scanning Electron Microscopy (SEM). Notably, SEM images revealed that antibodies specific for EV-associated biomarkers captured particles in the size range of 50-300 nm, with the majority smaller than 200 nm. In contrast, antibodies specific for soluble biomarkers did not capture particles larger than the resolution limit of the SEM protocol (approximately 30 nm). Once again, these size measurements align with the anticipated average size of EVs and monomeric proteins.

Example 6. EV-associated Biomarkers Are Stable Under Various Conditions and to Tissue Contamination. Direct detection is critical for improved performance.

EV-associated biomarkers in the blood are stable when proper measures are taken to store the samples. To evaluate if the EV-associated biomarkers isolated via methods described herein are stable, 100 uL aliquots of human plasma from two different individuals were isolated, subjected to 20 freeze-thaw cycles, and the CD9 and Amyloid beta associated EV complexes were measured using an EV-associated specific assays (FIG. 12A). The freeze-thaw cycles were performed by freezing the samples at -80 °C followed by thawing to room temperature. Samples were suspended in PBS alone or in combination with either 0.2 % Triton™ X-100 or 2% Triton™ X-100 to cause cell lysis (FIG. 12B). The detergent (i.e., surfactant) was used to evaluate if standard protocols for cell treatments impacted the integrity of the EV-associated biomarkers. There were small, insignificant variations in the marker quantity across each sample tested, including the samples treated with Triton™ X-100. These results show that EV-associated biomarkers in blood are stable to multiple types of common handling and treatment conditions. The remarkable stability of EV-associated biomarkers may be attributed to the inherent resilience of EVs to these conditions, which is then imparted to the biomarkers.

The inherent of resilience of EVs to common lysis conditions results in a specific phenomenon where the common procedure of sequentially isolating EVs, exposing them to a lysis solution, and measuring the lysate/supernatant results in a much lower signal than directly measuring the isolated EVs and their associated cargo (FIG. 11). Without being bound by any theory, it is hypothesized that this is because the EVs are resistant to common lysis conditions, so little cargo is released into the supernatant, and most of the EV (and their cargo) are found on the surface that they are captured on.

This was demonstrated with two types of cargo, Tau and amyloid beta, found within and on the surface of EVs. EVs that were treated even briefly with a mild non-polar (i.e., non-ionic) detergent (Tween® 20) had a much higher signal than untreated ones for the same cargo (FIG. 13). This shows that much of the tau and amyloid beta cargo associated with EVs are found internally within the EV. Furthermore, extended treatment with the detergent did not cause the signal to decrease (which would be expected if the EV was completely broken up), but actually caused a slight increase in signal (most likely from more permeabilization of the EV membrane). This further demonstrates how resilient the EVs are to lysis and why direct measurement of the EVs is giving a much stronger signal than the measurement of the lysate/supernatant. EVs are resilient to a wide range of commonly used detergents that have been reported to lyse EVs. The total EVs in a human blood sample were trapped on a surface using a CD81 antibody, treated with various detergents, and then the levels of tau protein left behind on the surface were measured. If EVs would be completely lysed, it would be expected that the signal be lower than the untreated EVs, as the tau protein cargo would be released into the lysate/supernatant. Instead, over a 100 times stronger signal of tau protein was achieved on the surface post-detergent treatment (FIG. 14B). This shows that the treatment with a wide range of detergents is able to permeabilize the EV membrane to measure internal cargo, but also that the EVs captured on the surface are resistant to lysis such that most of the cargo is not released into the lysate/supernatant, but instead remains associated with the EV complex on the surface.

Each subject is known to have a different composition of blood, which results in subject specific matrix interferences, and varying technical artifacts. In order to perform direct detection of the EV and their associated cargo, a universal way to cancel out blood matrix effects is needed, while not interfering with specific signals. This was accomplished by developing a sample diluent to premix all samples in that will specifically target and neutralize different potential interfering components. These components are defined and their effects are detailed in FIG. 10.

The specificity of the EV-associated biomarker against the soluble marker was assessed by measuring the signal when exposed to peripheral tissue contamination (FIG. 15A). The measured concentrations were normalized to reference standards. The neural EV-associated alpha synuclein and soluble alpha synuclein signals from two subject plasma samples were measured after spiking the samples with red blood cells (RBCs) treated with different amounts of freeze-thaw cycles from the same respective subject, yielding samples with intact RBCs, semi-lysed RBCs, and lysed RBCs. The EV-associated alpha synuclein signal remained stable across the three spiked samples for both subjects, whereas the soluble alpha synuclein signal increased as the amount of lysed RBCs increased in both subjects’ samples.

In a separate study, the stability of the EV-associated biomarker against the soluble marker was assessed by measuring the signal after increasing incubations times at room temperature before processing the plasma (FIG. 15B). The whole blood of six subjects was collected in K2EDTA blood collection tubes and were left at room temperature for up to 24 hours before the measurement. The signals from soluble alpha synuclein for each of the six subjects increased with incubation times, whereas the signals from the EV- associated alpha synuclein remained stable over 24 hours. These studies demonstrate that trace amounts of RBC lysis can cause significant alpha synuclein leakage into the sample, which neural EV-associated biomarkers are robust to.

Example 7. The Diagnostic Value of EV-associated Biomarkers Normalized to the Level of Total EVs

Sample and assay preparation for the below studies is described in Example 1 in line with the first generation assay. The methodology for obtaining the composite values and disease classifiers (or classification models) for the below studies are described in Example 10.

Amyloid Beta Aggregation in the Brain

One of the earliest pathological hallmarks of AD is brain deposits of Amyloid-beta. These plaques are formed from clustering abnormal amyloid protein fragments, primarily the hydrophobic variant Amyloidbeta peptide 42. Amyloid-beta proteins are released into the extracellular space and can circulate through the bloodstream. Also found in the extracellular space, exosomes are nanoscale membrane vesicles secreted by mammalian cells by fusing multivesicular endosomes with the plasma membrane. During this EV (e.g., exosome, microvesicle, apoptotic bodies) biogenesis, glycoproteins and glycolipids are incorporated into the invaginating plasma membrane and sorted into the newly formed exosomes. Through these surface markers, exosomes can associate with extracellular Amyloid-beta proteins. To evaluate EV-associated Amyloid beta in human blood plasma, 20 plasma samples, including amyloid positive and amyloid negative subjects determined by Amyloid-PET, were analyzed using an EV- associated Amyloid beta specific assay (FIG. 25C). The composite amyloid positive values were enriched against the composite amyloid negative values and yielded an accurate amyloid positivity classifier (AUC:0.870). The results of the discovery cohort were compared to the validation cohort which included 15 human plasma samples with amyloid positive and amyloid negative subjects, and similar performance was observed (AUC: 0.840) (FIG. 26).

Initially, the absolute concentrations of the EV-associated Amyloid beta 42, EV-associated GFAP, EV-associated NfL, and total EV signal per unit blood were calculated for both amyloid positive and healthy controls. For each disease marker, there was no significant difference between the absolute concentrations of the amyloid positive and healthy controls (FIG. 25A). After utilizing algorithmic correction to adjust for EV quantity bias (FIG. 25B) and calculating the composite values, EV quantity corrected (e.g., normalized) EV- associated Amyloid beta and amyloid beta 42 yielded an improved performance amyloid positivity classifier (AUC: 0.870) (FIG. 25C). This result demonstrates the ability to develop an amyloid positivity classifier from either the composite values of the individual EV-associated biomarkers or the combination of the composite values from different EV-associated biomarkers utilizing algorithmic correction. Notably, the latter provides an improved classification model for amyloid positivity diagnosis when utilizing both EV-associated Amyloid beta and Amyloid beta 42.

Brain Vascular Damage

Neuron-specific enolase (NSE) levels in a patient samples have gernerated great interest in recent years, resulting from extensive studies focusing on the role of NSE expression in CFS and its relationship to brain vascular damage, such as a stroke. The CNS cellular response to stroke results in characteristic upregulation and release of particular neuronal markers into the CSF and bloodstream. At present, there are few diagnostic methods for the assessment of a patient at risk for or undergoing a stroke. Given that NSE is found in the cytoplasm of neurons and cells with neuroendocrine differentiation, it will likely be found in EVs.

Another target of interest, SWOB, has gained notoriety in biomarker detection of brain vascular damage. SWOB is commonly known as a protective agent in cells. However, elevated levels of SWOB in the extracellular space may lead to cell damage and may be involved in neurodegenerative processes, such as stroke.

To evaluate the effectiveness of the EV-associated NSE and SWOB for classifying stroke, immunoassays for EV-associated SWOB and NSE were utilized. Specifically, a cohort of 45 patients containing ischemic stoke patients and healthy controls was evaluated. No significant difference was observed between the stroke and healthy controls for the absolute concentration of the EV-associated NSE, EV-associated S100B, and total EVs per unit volume of the blood (FIG. 27A). In utilizing algorithmic correction for the EV-associated NSE and SWOB (FIG. 27B) and then combining them to create a composite score, an accurate stroke classifier was produced (AUC: 0.8 W) (FIGs. 27C, 28A, and 32A). Conversely, the soluble forms of NSE and S100B did not yield an accurate stroke classifier (AUC: 0.572) (FIG. 28B). In this study, the level of biomarkers were not normalized to the level of total EVs. The EV-associated GFAP and NfL also did not yield an accurate stroke classifier (AUC: 0.506) (FIG. 28C). These results were replicated in a cohort of 50 patients containing ischemic and hemorrhagic stroke patients and healthy controls. An accurate stroke classifier was obtained for EV-associated NSE and S100B (AUC: 0.803) (FIGs. 29C and 33A), whereas the soluble forms of NSE and SWOB did not yield an accurate stroke classifier (AUC: 0.526) (FIG. 29B). In this study, the level of biomarkers were not normalized to the level of total EVs. Once again, the EV-associated GFAP and NfL also did not yield an accurate stroke classifier (AUC: 0.562) (FIG. 29C). This demonstrates that the specific cargo associated with the EV impart important functional information that can be used to reflect disease processes and accurately diagnose disease.

Alpha Synuclein Aggregation in the Brain

Parkinson’s disease is another neurological condition relating to alpha synuclein aggregation in the brain that may go undiagnosed in patients with little to no symptoms. Recently, work has been conducted to identify biomarkers associated with Parkinson’s disease to provide better preventative. Of special interest is the role alpha-synuclein plays in the pathology of Parkinson’s disease. Alpha-synuclein is upregulated in several conditions, including Parkinson’s disease and Alzheimer’s. The distribution of the pathology at the cellular and regional level is different in each disease. Thus, understanding the early presence of increased levels of alpha-synuclein is of great interest.

Additionally, GFAP is a second biomarker identified to be associated with the progression of neurological diseases such as Parkinson’s disease. GFAP is an astrocyte-specific intermediate filament protein that plays a crucial role in maintaining astrocyte's structural integrity and functioning. GFAP can diffuse into the cerebrospinal fluid during neuroinflammation and then enter the blood. This dissociation into the CSF and blood may be used for diagnosing a patient suspected of having Parkinson’s disease . Additionally, researchers sought to determine whether plasma GFAP could serve as a predictive marker for the conversion of Parkinson’s disease patients with cognitive impairment to dementia.

To evaluate the effectiveness of the EV-associated alpha-synuclein, phosphorylated alpha- synuclein, and GFAP for classifying Parkinson’s disease or the risk of developing Parkinson’s disease, a set of immunoassays for the EV-associated and soluble forms were utilized. Specifically, a cohort of 40 patients including patients diagnosed with Parkinson’s disease and healthy patients was evaluated, and the levels of EV-associated and soluble forms of alpha-synuclein, phosphorylated alpha-synuclein, and GFAP were measured. No significant difference was observed between the stroke and healthy controls for the absolute concentration of the EV-associated alpha-synuclein, EV-associated phosphorylated alpha-synuclein, EV- associated GFAP, EV-associated NfL, and total EVs per unit volume of the blood (FIG. 34A). In utilizing algorithmic correction for the EV-associated alpha-synuclein, phosphorylated alpha-synuclein, and GFAP, (FIG. 34B) then combining them to create a composite score, an accurate stroke classifier was produced (AUC: 0.866) (FIGs. 34C and 35A). Conversely, the soluble forms of alpha-synuclein, phosphorylated alpha- synuclein, and GFAP did not yield an accurate Parkinson’s disease classifier (AUC: 0.515) (FIG. 35B). In this study, the level of biomarkers were not normalized to the level of total EVs. EV-associated GFAP and NfL also did not yield an accurate Parkinson’s disease classifier (AUC: 0.539) (FIG. 35C).

These results were replicated in a cohort of 50 patients containing individuals with Parkinson’s Disease and healthy controls. EV-associated alpha-synuclein, phosphorylated alpha-synuclein, and GFAP yielded an accurate Parkinson’s disease classifier (AUC: 0.814) (FIG. 36A). The soluble forms of alpha- synuclein, phosphorylated alpha-synuclein, and GFAP did not yield an accurate Parkinson’s disease classifier (AUC: 0.596) (FIG. 36B). In this study, the level of biomarkers were not normalized to the level of total EVs. EV-associated GFAP and NfL did not yield an accurate Parkinson’s disease classifier (AUC: 0.518) (FIG. 36C). This demonstrates that the specific cargo associated with the EV impart important functional information that can be used to reflect disease processes and accurately diagnose disease.

Example 8. Selective Binding and Detection Utilizing EV Subpopulations

The potential diagnostic and prognostic applications of monitoring disease-induced modifications in EV cargo are considerable. By scrutinizing the molecular contents of EVs from different types of cells, especially those released under pathological conditions, researchers and scientists can gain a valuable window into the underlying processes associated with disease. This nuanced understanding opens avenues for more accurate diagnostic tools and prognostic indicators compared to the total EV-associated biomarkers.

Specifically, assays were initially designed to be sensitive to EV-associated biomarker epitopes for endogenous circulating biomarkers using both endogenous EVs and exogenous EVs. A human sample comprising at least two biomarkers was assessed using either only endogenous EVs or via incubating the sample with exogenous EVs via the methods described above (FIG. 16). Following incubation of the exogenous EVs, measurements using the assays described herein were performed, and compared to a healthy control sample. The measured concentrations were normalized to the total EVs within the EV type. It may be observed that selective binding occurs based on the EVs and biomarker properties. Specifically, a disease sample may have a higher concentration of EV type 1 biomarker binding. In contrast, the healthy sample may have a higher binding of a biomarker to the EV type 2, which may be an exogenously loaded EV. Thus, assays to measure how the same biomarker state may bind differently to different types of EVs based upon the EV subtype properties may be designed using binding agents that are specific for different EV subtypes. Different EV subtypes include endogenous or exogenous EVs from different cell types (EV subpopulation).

A cohort of thirty (30) patient samples was evaluated for Tau protein to evaluate the validity of the binding biomarkers' selectivity. Specifically, utilizing distinctive extracellular vesicle (EV) subpopulation markers, including CD9, CD171 , CD56, NRCAM, and GLAST, Tau binding to each type of EV in blood was evaluated. The measured concentrations were normalized to a reference EV standard. The expression of associated Tau was variable among the different subpopulations (FIG. 17). Analysis of the thirty (30) patient samples unveiled many patterns in the binding of Tau to different EV subpopulations across the samples. This variability underscores the biological diversity in the levels of the same biomarker across various brain- derived EVs, which is different from the total EV-associated biomarker signal. Without being bound by any theory, it is hypothesized that this variation could stem from differences in the local microenvironment at EV biogenesis or the production of EVs and the biomarker (Tau) in peripheral tissues (e.g., skeletal muscle), potentially confounding the signal of total EV-associated Tau. These results also underscore how the EV subpopulation may provide diagnostic value, as different brain-derived EVs, as compared to the total EVs, also contain different levels of the same biomarker that is biologically different. Example 9. The Diagnostic Value of EV-associated Biomarkers Normalized to the Level Total EVs within an Isolated Subpopulation

Sample and assay preparation for the below studies is described in Example 1 in line with the second generation assay. The methodology for obtaining the composite values and disease classifiers (or classification models) for the below studies are described in Example 10.

Brain Vascular Damage

Without being limited by any theory, it may be hypothesized that looking at only specific EV subpopulations would identify and remove noise from the biologically meaningful signals. To prove that the EV subtype-bound biomarker would improve disease classification accuracy for diseases relating to brain vascular damage, the CD171 (Brain EV), CD56 (Neural lineage EV), NrCAM (Neuron EV), GLAST (Astrocyte EV), CD9 (Pan EV), CD81 (Pan EV) bound NSE and SWOB levels in blood plasma were measured with a set of immunoassays in a cohort of 28 subjects containing ischemic stroke patients and healthy age-matched controls. Initially, the amounts of brain EV-associated NSE & brain EV-associated SWOB, astrocyte EV-associated NSE & astrocyte EV-associated SWOB, were normalized against total brain EVs and astrocyte EVs counts respectively were calculated for both stroke and healthy controls. For each disease marker, there was no significant difference between the absolute concentrations of the stroke and healthy controls (FIG. 30A). Upon correction for respective EV subpopulation quantities, there was better separation between the stroke and healthy controls (FIG. 30B) However, it was the combination of all four markers into a composite score that best classified stroke and healthy samples (FIGs. 30C and 32B) These results demonstrate the ability to develop an accurate stroke classifier from either the composite values of the individual EV-associated biomarkers or the combination of the composite values from different EV- associated biomarkers by way of algorithmic correction. The result also demonstrates the ability to develop an accurate stroke classifer using EV-associated biomarkers from specific EV subpopulations.

For the same cohort, amounts of brain, astrocyte, and neural EV normalized to total EVs (CD81- CD9) per unit volume composite values were calculated for both stroke and healthy controls. For each disease marker, there was no significant difference between the absolute concentrations of the stroke and healthy controls (FIG. 31 A). The brain, astrocyte, and neuronal EVs, however, yielded an accurate stroke classifier after correcting each EV subpopulation against total EV quantity (FIG. 31 B) and calculating the composite values (AUC: 0.988) (FIGs. 31 C and 32C). This result demonstrates the ability to develop a stroke diagnostic solely using EV-markers characteristic of EV subpopulations as the biomarkers of disease. It is hypothesized that this model captures the differences in brain cell health as a result of blood-brain barrier damage in the acute ischemic stroke event that causes changes in EV production level of each cell type.

These results are compared to the stroke classifier obtained for the total EV-associated NSE and SWOB within the cohort. The total EV-associated NSE and SWOB were able to classify ischemic stroke patients and healthy controls (AUC: 0.815) (FIG. 32A, 28A, and 27C). However, the performance was vastly improved when isolating the signal to only brain and astrocyte EV-associated NSE and SWOB (FIG. 32B and 30C) and when isolating the signal to only brain, neuron, and astrocyte EV (FIG. 32C and 31 C). It is hypothesized that these neuron and astrocyte EV-associated models capture the specific signal measuring brain cell health in contrast to the total EV-associated model which potentially also contain NSE and SWOB originating from other tissues, adding noise to the signal.

These results were replicated in the second independently collected cohort of 50 subjects containing

5 ischemic and hemorrhagic stroke patients and healthy age-matched controls previously described in Example 7. As previously shown, the total EV-associated NSE and SWOB were able to accurately classify ischemic stroke patients and healthy controls (AUC: 0.803) (FIGs. 33A and 29A). When isolating the signal to only the brain and astrocyte bound NSE and SWOB and ignoring the signal from neuron EV-associated and total EV-associated forms of the same biomarkers, the performance greatly improved (AUC: 0.940)

W (FIG. 33B). Furthermore, looking at the levels of brain, neuron, and astrocyte EVs, a classification model with greatly improved performance was produced (AUC: 0.924) (FIG. 33C).

Alpha Synuclein Aggregation in the Brain

To prove that the EV subtype-bound biomarker would improve disease classification accuracy, the CD171 (Brain EV), CD56 (Neural lineage EV), NrCAM (Neuron EV), GLAST (Astrocyte EV), CD9 (Pan EV),

15 CD81 (Pan EV) bound alpha-synuclein, phosphorylated alpha-synuclein, and GFAP levels in blood plasma were measured with a set of immunoassays in a cohort of 26 patients containing individuals with Parkinson’s disease and healthy controls.

The total EV-associated alpha-synuclein, phosphorylated alpha-synuclein, and GFAP were not significantly different between PD and healthy controls (FIG. 37A), but after correcting for EV quantity (FIG.

20 37B) and combining them into a composite score, it yielded an accurate Parkinson’s disease classifier (AUC: 0.839) (FIGs. 37C and 40A). In observing the EV-associated alpha-synuclein and phosphorylated alpha- synuclein isolated to brain EV subpopulations, a similar trend was observed with raw quantities (FIG. 38A), corrected values (FIG. 38B), and the composite score performance of the Parkinson’s disease classifier vastly improved (AUC: 0.976) (FIGs. 38C and 40B). Moreso, in observing the EV-associated alpha-synuclein

25 and phosphorylated alpha-synuclein isolated to the neuron and astrocyte EV subpopulation, the composite score combining all these measure also yielded a vastly improved Parkinson’s disease classifier (AUC: 0.964) (FIGs. 39C and 40C). It is hypothesized that these brain, neuron, and astrocyte EV-associated models capture the specific signal measuring brain cell health in contrast to the total EV-associated model which potentially also contain alpha-synuclein, phosphorylated alpha-synuclein, and GFAP originating from

30 other tissues, adding noise to the signal. It is further hypothesized that GFAP improved the total EV associated biomarker signature but not the brain, neuron, and astrocyte EV signatures because it added some information about brain specificity to the total EVs that was not needed once brain derived EV markers were used.

Tau Aggregation in the Brain

35 Alzheimer's disease is marked by the accumulation and aggregation of misfolded proteins in the brain, notably Beta-Amyloid and Tau, leading to cognitive decline. It is believed that these two types of protein aggregates may interact synergistically to promote neuronal dysfunction and neurodegeneration in Alzheimer's disease. Tau normally supports neuronal structure, but in Alzheimer's, it becomes hyperphosphorylated, misfolding and forming tangles within neurons, disrupting their function and contributing to their degeneration.

Although Tau proteins are typically intracellular, they can be released into the extracellular space. Their interaction with extracellular vesicles (EVs), facilitates their transport within the brain and even beyond, crossing the blood-brain barrier into the circulation.

To prove that the EV subtype-bound biomarker would improve disease classification accuracy, the CD171 (Brain EV), CD56 (Neural lineage EV), NrCAM (Neuron EV), EAAT1 (Astrocyte EV), CD9 (Pan EV) , CD81 (Pan EV) bound Tau and phosphorylated Tau (pTau) levels in blood plasma were measured with a set of immunoassays in a cohort of 34 subjects consisting of 16 Tau-PET positive patients and 18 healthy age- matched controls. Unlike the brain vascular damage indication, the total EV-associated Tau and pTau was unable to differentiate the Tau-PET positive subjects from the healthy controls (AUC: 0.590) (FIG. 41C). However, when the signal was isolated to only the brain and astrocyte EV-associated forms of Tau and pTau, an accurate tau aggregation in the brain classifier was produced (AUC: 0.835) (FIG. 42C and 44B). A Tau-PET positive classifier using a composite of neuron EV-associated forms of Tau and pTau also achieved good performance (AUC: 0.900) (FIG. 43C and 44C). It is hypothesized that these brain, neuron and astrocyte EV-associated models capture the specific signal measuring brain cell health in contrast to the total EV-associated model which potentially also contain Tau originating from many other tissues, adding noise to the signal.

This cohort was expanded with an additional independently collected 14 healthy subjects, 2 Tau- PET positive, Amyloid-PET negative subjects, and 4 Tau-PET negative, Amyloid-PET positive subjects. Amyloid-PET positive individuals were included in the study to demonstrate the selectivity of the neuronal EV-associated immunoassays toward tau and or amyloid brain aggregation state.

Using the original trained blood biomarker signature using a composite of neuron EV-associated Tau and pTau and a cutoff that gave 91% accuracy, 13/14 healthy controls were correctly classified, 2/2 Tau-PET positive, Amyloid-PET negative samples were correctly classified, and 4/4 Tau-PET negative, Amyloid-PET positive samples were correctly classified (FIG. 45A). This shows that neuronal EV-associated forms of tau cargo are specific to tau aggregation in the brain and not amyloid aggregation. The performance of the logistic regression classifier (with leave-out-out cross validation) on the original 16 Tau-PET positive subjects and all 32 healthy samples in the expanded cohort. Its performance (AUC: 0.936) (FIG. 45B) is comparable to the previous model (FIG. 44C, AUC: 0.917), showing that the classifier is specific to tau pathology and not to sample collection or site differences. In testing this trained classifier (or classification model) with the Tau- PET and Amyloid-PET typed samples as input, the Tau-PET positive samples had a higher Tau aggregation prediction score than the Tau-PET negative samples, despite Amyloid-PET positivity. However, Tau-PET positive with Amyloid-PET positivity samples demonstrated higher prediction scores on average compared to Tau-PET positive with Amyloid-PET negativity samples. Without being bound by any theory, it is hypothesized that the neuronal EV-associated phosphorylated Tau is specifically diagnostic of Tau aggregation and not Amyloid aggregation in the brain due to the local environment of EV biogenesis. In contrast, soluble pTau 217 measured in blood samples has been shown to be more specific to brain amyloid aggregation. TDP43 aggregation

TDP43 aggregation is a hallmark of several neurodegenerative diseases (or, e.g., TDP43 proteinopathies), including amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). In healthy cells, TDP43 is involved in RNA processing and gene regulation. However, in disease states, it misfolds and forms abnormal aggregates in the cytoplasm, leading to a loss of its normal functions and contributing to neurodegeneration. These aggregates disrupt cellular homeostasis, induce cellular stress, and can trigger inflammatory responses, ultimately leading to neuronal cell death and the progression of these debilitating conditions. Similar to prions, misfolded TDP43 can induce conformational changes in neighboring, normally folded TDP-43 proteins, leading to further aggregation. This self-propagating mechanism allows TDP43 aggregates to spread trans-synaptically across neuronal networks, contributing to the progressive nature of diseases like ALS and frontotemporal dementia. The spread of TDP-43 aggregates is thought to involve mechanisms such as EV release and uptake by neighboring cells, perpetuating the cycle of misfolding and cellular dysfunction in the CNS. This misfolded TDP43 packaged in and on CNS EVs facilitates its transport within the brain and even beyond, crossing the blood-brain barrier into the circulation.

To prove that the EV subtype-bound biomarker would improve disease classification accuracy, the CD171 (Brain EV), CD56 (Neural lineage EV), NrCAM (Neuron EV), GLAST (Astrocyte EV), SYP (synapse), CD81 (total EV) bound TDP43 and phosphorylated TDP43 (pTDP43) levels in blood plasma were measured with a set of immunoassays in a cohort of 36 subjects consisting of 13 ALS subjects and 23 healthy age- matched controls. Like the tau aggregation results (AUC: 0.590) (FIG. 41 C), the total EV-associated TDP43 and pTDP43 were unable to differentiate the ALS subjects from the healthy controls (AUC: 0.599) (FIG. 47C). However, when the signal was isolated to only the neuron and neuronal EV-associated forms of TDP43 and pTDP43, an accurate ALS classifier was produced (AUC: 0.893) (FIG. 48C). It is hypothesized that these neuron and neuronal EV-associated models capture the specific signal measuring brain cell health in contrast to the total EV-associated model which potentially also contain TDP43 and pTDP43 originating from many other tissues, adding noise to the signal.

Example 10. Composite EV-associated Biomarker Value Via Normalization Against Level of Total EVs or Total EVs Within a Subpopulation

Immunoassays were developed to measure the total EV quantity (e.g., count or level) in blood samples across seven different clinical cohorts, yielding a total EV quantity range of over 4 orders of magnitude across the three cohorts (FIG. 22A). For randomly selected samples in the same cohorts, the EV- associated biomarker quantity for tau, phosphorylated tau, Amyloid beta, Amyloid beta 42, GFAP, and NfL were also measured using immunoassays specific for each type of EV-associated biomarker (FIG. 22B). Notably, the EV-associated biomarker quantity for each biomarker positively correlated with the total EV quantity for every subject.

In the experiment showing consistency between alternate configurations of antibodies (FIG. 7), additional assays were designed to measure the total EV count by using the binding agent specific toward different EV associated biomarkers as both the capture and detection agent (CD56.CD81 for total neural lineage EV count and CD81 .CD81 for total EV count). In comparing the level of total EV count and total neural lineage EV count against the level of EV-associated biomarker for each type of EV marker, the total EV count generally corelated with the level of EV-associated Tau and EV-associated alpha synuclein (FIG. 23). However, a few samples deviated from the general trendline, while these same samples did not deviate when tested for specificity to its alternate configurations. Without being bound by any theory, it is hypothesized that other factors such as the individual, type of cargo, and type of EV result in these deviations.

Accordingly, a measure of the quantity of an EV-associated biomarker may not classify disease against a control, as the total EV quantity in the blood depends on many factors and is not purely dependent on disease state. Such bias in the measurement can be overcome by generating a compositive value representative of the level of EV-associated biomarker and the level of total EVs in the sample. For example, the composite value can be calculated by regressing the EV-associated biomarker quantity against the total EV quantity (FIGs. 24A and 42B). The EV-associated amyloid beta 42 quantity positively correlates with the EV quantity in blood of both amyloid negative and positive groups. However, there is no classification between the positive and negative controls. Once the EV-associated amyloid beta 42 quantity (or level) is normalized against the EV quantity (regression adjustment), the bias is corrected for, providing separation between the amyloid positive and negative controls.

To correct for the EV quantity confounding effects, a series of regression models were performed to regress out the influence of total EV levels. The composite values were calculated for EV-associated biomarkers in Examples 7 and 9 for both disease and healthy controls unless otherwise specified. When the total EV-associated biomarker is measured within a sample, composite values were calculated for each EV- associated biomarker by normalizing the level of EV-associated biomarker against the level of total EVs within the sample. When the total EV-associated biomarker was measured within an EV subpopulation, composite values were calculated for each EV-associated biomarker by normalizing the level of EV- associated biomarker against the level of total EVs within the corresponding EV subpopulation. For example, the astrocyte EV-associated alpha synuclein composite value was calculated by normalizing the level of astrocyte EV-associated alpha synuclein against the level of total astrocyte EVs (e.g., by utilizing a GLAST.CD9 or GLAST.CD81 antibody pair). The composite values are, thus, also referred to as the normalized values or normalized signals.

Assays for determining the total level of EVs within the sample utilized a CD9.CD9, CD81 .CD9, CD9.CD81 , CD81 .CD81 antibody pair. Assays for determining the total level of EVs within the corresponding EV subpopulation utilized an antibody pair including an antibody that preferentially binds to either CD9 or CD81 and an antibody that preferentially binds to an EV-marker specific to the EV subpopulation (or cell type).

All potential combinations of measured levels may be utilized to produce a composite value of diagnostic utility. For example, composite values can also include the levels of two different EV-associated biomarkers or the levels of two different EV markers. In a second example, the level of total EVs or the level of total EVs within an EV subpopulation of a specific cellular origin may be normalized against the level of EV-associated disease biomarker. Additional types of normalization are detailed in Examples 12-18. Example 11. Algorithmic Combination and Correction of Composite Values

To produce an accurate classifier for disease, the composite values for one or more biomarker were algorithmically corrected and combined for the above Examples 7 and 9, unless otherwise specified. When composite values were not calculated, the absolute concentration of one or more biomarker were algorithmically corrected and combined for the above Examples 7 and 9. Algorithmic correction was performed by constructing multiple logistic regression classifiers with leave-one-out cross validation on each corrected EV-associated biomarker value (e.g., level) or absolute concentration of unbound biomarker. Multiple logistic regression models were constructed to assess the relationship between the independent variables and the outcome while controlling for identified confounders. The model's goodness-of-fit was evaluated using leave-one-out or five-fold cross-validation of the datasets. Logistic regression assumptions were assessed and fulfilled, including linearity, independence of errors, and absence of multicollinearity. The significance level for all analyses was set at a = 0.05 after Bonferroni multiple hypothesis correction. All statistical analyses were performed using R v4.3.2 and GraphPad Prism 10. The classification model derived from the algorithmic correction was then used to provide prediction scores for the disease and healthy controls. This algorithmic correction also results in a classifier that combines all EV-associated biomarkers measured. The predicted probability is then calculated utilizing the following equation:

where x_n is the weight determined for each composite value utilizing the algorithm.

Finally, feature selection algorithms were used to remove non-informative biomarkers from the classification model. The feature selection algorithms utilized include forward selection, recursive feature elimination (e.g., fuzzy recursive feature elimination), and penalized regression algorithms.

As shown in the previous Examples, the absolute concentrations and/or composite values can yield accurate classifiers of disease for particular EV-associated biomarkers. However, EV-associated biomarkers must be surveyed for their relevance toward disease diagnosis and contribution toward developing an accurate disease classifier. EV markers were assessed for their relevance toward stroke diagnosis within three clinical cohorts having containing stroke patients and healthy controls (see Example 9). Although literature indicated the potential diagnostic relevance of NSE and SWOB for stroke diagnosis, there was still substantial overlap between the two populations for the EV-associated biomarkers. In utilizing algorithmic correction, as was done in FIGs. 27A-33C, accurate stroke classifiers were obtained from EV-associated NSE and SWOB among total EVs and within EV subpopulations. In the latter, the classification model also utilized the brain EV-associated NSE composite values which did not, alone, yield an accurate stroke classifier (FIGs. 30B and 31 B), demonstrating how an otherwise insignificant EV-associated biomarker can lead to an improved disease classifier once combined with other EV-associated biomarkers. These results indicate the importance of utilizing algorithmic correction, along with the feature selection algorithms, to extract accurate disease classifiers. Example 12. Resulting Diagnostic Value of Various Composite Values and Their Two-Way and Three- Way Combinations toward Diagnosis of Amyloid Beta Aggregation in the Brain Utilizing First Generation Assays

The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 1 Assay Design in FIG. 2) to determine their diagnostic value toward determining amyloid beta aggregation in the brain of a subject: amyloid beta (Ap), amyloid beta 42 peptide (Ap 42), GFAP, Tau, NfL, and CD9. The EV-associated biomarkers were measured from blood plasma from 20 plasma samples, including amyloid positive (disease) and amyloid negative (healthy) subjects as determined by Amyloid-PET. Results from the EV-associated Ap, Ap 42, EV-associated GFAP, and EV-associated CD9 assays are shown in FIGs. 25A-25C, 26A, and 26B. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 2. Accuracy of EV-associated Biomarker Levels without Normalization.

Table 3. Accuracy of Single Composite Values.

Table 4. Accuracy of Combination of Two Composite Values.

Table 5. Accuracy of Combining Three Composite Values.

The markers in Table 2 were utilized as inputs for composite values in Tables 3-5. Table 2, not utilizing normalization or algorithmic combination and correction, yields a maximum AUC of 0.62. However, with normalization, as shown in Table 3, improved amyloid aggregation classifiers were obtained (highest AUC: 0.88). Algorithmic correction and combination of two composite values yielded higher AUCs than a single composite value, as shown in Table 4 (highest AUC: 0.93). Table 5 shows that combination of three composite values yielded combinations with an AUC of 0.96, demonstrating how additional composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed. Various composite value components were tested, including, e.g., the level of EV-associated disease biomarkers (e.g., Ap 42) normalized to total EVs (e.g., CD9) and the level of one EV-associated disease biomarker (e.g., Ap 42) normalized to the level of a second EV-associated disease biomarker (e.g., Tau).

Example 13. Resulting Diagnostic Value of Various Composite Values and Their Two-Way and Three- Way Combinations toward Brain Vascular Damage Utilizing First Generation Assays

The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 1 Assay Design in FIG. 2) to determine their diagnostic value toward determining brain vascular damage in a subject: GFAP, NfL, NSE, S100B, and CD9. The EV-associated biomarkers were measured from blood plasma from a cohort of patients containing ischemic stoke patients and healthy controls. Results from the EV-associated NSE, SWOB, and CD9 assays are shown in FIGs. 27A-29C. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 6. Accuracy of EV-associated Biomarker Levels without Normalization.

| CD9 | 0.57 | 0.54 |

Table 7. Accuracy of Single Composite Value.

Table 8. Accuracy of Combination of Two Composite Values.

Table 9. Accuracy of Combination of Three Composite Values.

The markers in Table 6 were utilized as inputs for composite values in Tables 7-9. Table 6, not utilizing normalization or algorithmic combination and correction, yields a maximum AUC of 0.61 . However, with normalization, as shown in Table 7, improved brain vascular damage classifiers were obtained (highest AUC: 0.74). Algorithmic correction and combination of two composite values yielded higher AUCs than a single composite value, as shown in Table 8 (highest AUC: 0.82). Table 9 shows that combination of three composite values yielded combinations with an AUC of 0.76, demonstrating how algorithmically corrected and combined composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed. Various composite value components were tested, including, e.g., the level of EV-associated disease biomarkers (e.g., NSE) normalized to total EVs (e.g., CD9) and the level of one EV-associated disease biomarker (e.g., NSE) normalized to the level of a second EV-associated disease biomarker (e.g., S100B).

Example 14. Resulting Diagnostic Value of Various Composite Values and Their Two-Way and Three- Way Combinations toward Alpha Synuclein Aggregation Utilizing First Generation Assays

The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 1 Assay Design in FIG. 2) to determine their diagnostic value toward determining alpha synuclein aggregation in the brain of a subject: alpha-synuclein (aSyn), phosphorylated alpha-synuclein (paSyn129), GFAP, NfL, and CD9. The EV-associated biomarkers were measured from blood plasma from a cohort of 40 patients including patients diagnosed with Parkinson’s disease and healthy patients. Results from the EV- associated aSyn, paSyn, GFAP, NfL, and CD9 assays are shown in FIGs. 34A-37C. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 10. Accuracy of EV-associated Biomarker Levels without Normalization.

Table 11 . Accuracy of Single Composite Value.

Table 12. Accuracy of Combination of Two Composite Values.

Table 13. Accuracy of Combination of Three Composite Values.

The markers in Table 10 were utilized as inputs for composite values in Tables 1 1 -13. Table 10, not utilizing normalization or algorithmic combination and correction, yields a maximum AUC of 0.61 . However, with normalization, as shown in Table 1 1 , improved alpha-synuclein aggregation classifiers were obtained (highest AUC: 0.72). Algorithmic correction and combination of two composite values yielded higher AUCs than a single composite value, as shown in Table 12 (highest AUC: 0.80). Table 13 shows that combination of three composite values yielded combinations with an AUC of 0.82, demonstrating how algorithmically corrected and combined composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed. Various composite value components were tested, including, e.g., the level of EV-associated disease biomarkers (e.g., aSyn) normalized to total EVs (e.g., CD9) and the level of one EV-associated disease biomarker (e.g., aSyn) normalized to the level of a second EV-associated disease biomarker (e.g., paSyn129).

Example 15. Resulting Diagnostic Value of Various Composite Values and Their Two-Way, Three- Way, and Four-Way Combinations toward Brain Vascular Damage Utilizing Second Generation Assays

A set of immunoassays were designed to measure the level of EV-associated biomarkers bearing two types of EV cargo within a sample, along with measuring the level of total EVs and total EVs within an EV subpopulation, to determine their diagnostic value toward determining brain vascular damage. The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 2 Assay Design in FIG. 2): NSE and S100B. The following EV markers were measured using EVA assays to associate the biomarker with EVs (either total EVs or within an EV subpopulation): CD171 , CD56, NrCAM, GLAST, CD81 , and CD9. The EV-associated biomarkers were measured from blood plasma from a cohort of subjects containing ischemic stroke patients and healthy age-matched controls. Results from the EV- associated NSE, S100B, CD171 , CD56, NrCAM, GLAST, CD81 , and CD9 assays are shown in FIGs. 30A- 33C. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers and surface markers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 14. Accuracy of EV-associated Biomarker Levels without Normalization.

| CD171 .NSE | 0.41 | 0.41 |

Table 15. Accuracy of Single Composite Value.

Table 16. Accuracy of Combination of Two Composite Values.

Table 17. Accuracy of Combination of Three Composite Values.

Table 18. Accuracy of Combination of Four Composite Values.

The markers in Table 14 were utilized as inputs for composite values in Table 15. Similarly, Tables 16-18 utilized the markers in Table 14 as inputs for both, all three, or all four composite values, respectively, which were combined to assess their diagnostic value through algorithmic correction. Tables 16-18 each show the top 150 results in terms of the highest AUC values. As shown in Tables 17 and 18, the combination of three composite values yielded over 150 combinations with an AUC of 1 .00 and an average error of 0, demonstrating how additional composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed.

Example 16. Resulting Diagnostic Value of Various Composite Values and Their Two-Way and Three- Way Combinations toward Alpha Synuclein Aggregation in the Brain Utilizing Second Generation Assays

A set of immunoassays were designed to measure the level of EV-associated biomarkers bearing two types of EV cargo within a sample, along with measuring the level of total EVs and total EVs within an EV subpopulation, to determine their diagnostic value toward determining alpha synuclein aggregation in the brain. The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 2 Assay Design in FIG. 2): aSyn, paSyn129, and GFAP. The following EV markers were measured using EVA assays to associate the biomarker with EVs (either total EVs or within an EV subpopulation): CD171 , CD56, NrCAM, GLAST, CD81 , and CD9. The EV-associated biomarkers were measured from blood plasma from a cohort of 26 subjects with Parkinson’s Disease and healthy age-matched controls. Results from the EV-associated aSyn, paSyn129, GFAP, CD171 , CD56, NrCAM, GLAST, CD81 , and CD9 assays are shown in FIGs. 37-40. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers and surface markers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 19. Accuracy of EV-associated Biomarker Levels without Normalization.

Table 20. Accuracy of Single Composite Value.

Table 21 . Accuracy of Combination of Two Composite Values.

Table 22. Accuracy of Combination of Three Composite Values.

Table 23. Accuracy of Combination of Four Composite Values.

The markers in Table 19 were utilized as inputs for composite values in Table 20. Similarly, Table 21 -23 utilized the markers in Table 19 as inputs for both, all three, or all four composite values, respectively, which were combined to assess their diagnostic value through algorithmic correction. Table 21 shows the top 150 results (AUC > 0.84) in terms of the highest AUC values. As shown in Tables 22 and 23, the combination of three composite values yielded over 150 combinations with an AUC of 0.91 and above, demonstrating how additional composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed. Example 17. Resulting Diagnostic Value of Various Composite Values and Their Two-Way and Three- Way Combinations toward Tau Aggregation in the Brain Utilizing Second Generation Assays

A set of immunoassays were designed to measure the level of EV-associated biomarkers bearing two types of EV cargo within a sample, along with measuring the level of total EVs and total EVs within an EV subpopulation, to determine their diagnostic value toward determining tau aggregation in the brain. The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 2 Assay Design in FIG. 2): Tau, pTau396, and pTau231 . Phosphorylation is present at amino acid positions 231 & 396, corresponding to pTau231 and pTau 396, respectively. The following EV markers were measured using EVA assays to associate the biomarker with EVs (either total EVs or within an EV subpopulation): CD171 , CD56, NrCAM, GLAST, CD81 , and CD9. The EV-associated biomarkers were measured from blood plasma from a cohort of 34 subjects with Tau-PET positivity and healthy age-matched controls. Results from the EV-associated Tau, pTau396, pTau231 , CD171 , CD56, NrCAM, GLAST, CD81 , and CD9 assays are shown in FIGs. 41 A-44C. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers and surface markers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 24. Accuracy of EV-associated Biomarker Levels without Normalization.

Table 25. Accuracy of Single Composite Value.

Table 26. Accuracy of Combination of Two Composite Values.

Table 27. Accuracy of Combination of Three Composite Values.

Table 28. Accuracy of Combination of Four Composite Values.

The markers in Table 24 were utilized as inputs for composite values in Table 25. Similarly, Tables 26-28 utilized the markers in Table 24 as inputs for both, all three, or all four composite values, respectively, which were combined to assess their diagnostic value toward tau aggregation through algorithmic correction. Table 25 shows improved AUC for single composite values (highest AUC: 0.64) compared to not normalizing levels (Table 24, highest AUC: 0.86). Tables 26-28 show the top 150 results (AUC > 0.86) in terms of the highest AUC values, demonstrating how algorithmic combination and correction of composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed.

Example 18. Resulting Diagnostic Value of Various Composite Values and Their Two-Way, Three- Way, and Four-Way Combinations toward ALS Utilizing Second Generation Assays

A set of immunoassays were designed to measure the level of EV-associated biomarkers bearing two types of EV cargo within a sample, along with measuring the level of total EVs and total EVs within an EV subpopulation, to determine their diagnostic value toward determining tau aggregation in the brain. The following EV-associated biomarkers were measured using EVbA, UA, and/or EVA assays (see, e.g., Gen 2 Assay Design in FIG. 2): aSyn, paSyn (e.g., paSyn129), GFAP, NfL, TDP (e.g., TDP43), and pTDP (e.g., pTDP43). The following EV markers were measured using EVA assays to associate the biomarker with EVs (either total EVs or within an EV subpopulation): CD171 , CD56, NrCAM, GLAST, CD81 , and SYN. The EV- associated biomarkers were measured from blood plasma from a cohort of 36 subjects with ALS and healthy age-matched controls. Results from the EV-associated TDP43, pTDP43, CD171 , CD56, and CD81 assays are shown in FIGs. 47A-48C. The resulting measured levels of each biomarker among the healthy and disease controls were used to determine the diagnostic value of the biomarkers and surface markers by calculating the AUC. To observe if improved diagnostic value can be extracted from the data, composite values and their combination were utilized, the latter involving the use of algorithmic correction. Algorithmic combination and correction are detailed in Example 11 .

Table 29. Accuracy of EV-associated Biomarker Levels without Normalization.

Table 30. Accuracy of Single Composite Value.

Table 31 . Accuracy of Combination of Two Composite Values.

Table 32. Accuracy of Combination of Three Composite Values.

Table 33. Accuracy of Combination of Four Composite Values.

The markers in Table 29 were utilized as inputs for composite values in Table 30. Similarly, Tables 31 -33 utilized the markers in Table 29 as inputs for both, all three, or all four composite values, respectively, which were combined to assess their diagnostic value toward ALS through algorithmic correction. Table 30 shows the same maximum AUC for single composite values (highest AUC: 0.69) compared to Table 29 which shows the measured levels of each marker without normalization. Tables 31 -33 show the top 150 results (AUC > 0.75) in terms of the highest AUC values, demonstrating how algorithmic combination and correction of composite values can lead to improved diagnostic value. Composite combinations with lower AUC values, yet still providing diagnostic value, were also observed.

Example 19. Competitive Binding Between Endogenous and Exogenous EVs

To demonstrate competitive binding between endogenous and exogenous EVs, assays were initially designed to be sensitive to an EV-associated biomarker epitope for an endogenous circulating EV- associated biomarker using endogenous and exogenous EVs (FIG. 18). A human sample comprising endogenous EV-associated biomarkers for one type of biomarker was assessed before and after incubation with one type of exogenous EV and further incubation with a second type of exogenous EV. Measurements using the assays described herein were performed and compared to a healthy control sample. It may be further observed that selective binding occurs based on the EVs and biomarker properties. After incubation with the first type of exogenous EV, the EV-associated biomarker signal reduced more for the disease sample compared to the healthy sample, in both cases, compared the EV-associated biomarker signal for the endogenous EV. Upon incubation with the second type of exogenous EV, the EV-associated biomarker signal decreased for the disease sample, while the EV-associated biomarker signal decreased further for the healthy sample. Without being bound by any theory, it is hypothesized that competitive binding occurs between the endogenous EVs and the two types of exogenous EVs, such that a different type of EV can promote dissociation of the biomarker from the endogenous EV to bind to the different type of EV. Furthermore, it is hypothesized that, in binding to the different type of EV, the EV-associated biomarker may expose a different epitope. Finally, it is hypothesized that as the relative EV-associated biomarker signal changes differed between the disease sample and healthy sample, the local environment of EV biogenesis may influence the properties of the EVs and biomarker.

To further demonstrate the competitive binding between varied binding affinities between exogenous EV subpopulations and targets, two human subjects' plasma samples were incubated with equal concentrations of cell line-derived EVs (exogenous EVs) and incubated for a sufficient time to allow binding to the soluble biomarkers (Tau and Amyloid Beta). The endogenous EVs were assayed using assay configurations as described above, and the endogenous Tau and Amyloid Beta were evaluated (FIG. 19). The measured signals were normalized to an exogenous EV-associated Tau signal, where the maximum value of 1 indicates the same signal measured for exogenous EVs incubated with a highly concentrated solution of recombinant Tau protein. It was observed that the Tau expression in plasma sample A was varied depending upon the incubation of EVs from neuronal EVs, astrocyte EVs, or epithelial EVs. When compared to plasma sample B, it was observed that the concentration of Tau also varied based on exogenous EVs. The same variations were seen in the Amyloid Beta protein concentrations for patient A and Patient B samples. To explore how EVs from different cell types influence the binding properties between a biomarker and endogenous EVs, an assay was designed to be sensitive to an EV-associated Tau epitope using endogenous EVs. Two human plasma samples comprising endogenous EV-associated Tau were assessed before and after incubation with increasing doses of epithelial cancer EVs (A431 cell line), neuronal EVs (SHSY5Y cell line), cow blood isolated EVs, goat blood isolated EVs, and chicken blood isolated EVs. The signal was normalized against a standard (e.g., exogenous EV-bound Tau reference) such that the EV- associated Tau for endogenous EVs in each sample (untreated sample) was set to 1 . The untreated samples for both patients were treated with increasing levels of each exogenous EVs (0.01 , 0.1 , and 1 ). Between the two human plasma samples, the same dose of the same type of EV elicited different responses and increasing doses of the same type of EV elicited different slopes of response (FIGs. 20A-20C). In all cases, the EV-associated Tau signal decreased compared to the untreated sample as the sample was incubated with an increasing amount of the different type of EV. Without being bound by any theory, it is hypothesized that the biological form of endogenous Tau protein is different among different patients. Further, the results demonstrate that the presence of exogenous EVs within a human sample can negatively affect the EV-associated biomarker signal derived from endogenous EVs as a result of competitive binding between the endogenous EVs and exogenous EVs.

To evaluate this complex interaction of exogenous EVs with a blood sample, cell-line derived EVs produced by neuronal cells were spiked into plasma samples, allowing the Tau protein in the human subject’s plasma to associate with them. Human blood samples were incubated with neuronal cell line- derived EVs (exogenous), and the levels of Tau bound to the exogenous EVs was evaluated (FIGs. 21 A and 21 B). Between 10⁶ and 10⁹ EVs were incubated in the blood samples for up to 4 hours at room temperature, and a biomarker assay was used to measure the association of Tau protein to the EVs. The signals were normalized to an exogenous EV-associated Tau reference. The signal increased proportional to the number of exogenous EVs spiked into the human sample (FIG. 21A). Additionally, the Tau signal also increased proportional to the incubation time of the EVs (FIG. 21 B). The concentration and time-dependent increases observed supported the hypothesis that these EVs may act as a scaffold for the biomarkers present in the sample.

Other Embodiments

While the invention has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the invention that come within known or customary practice within the art to which the invention pertains and may be applied to the essential features hereinbefore set forth, and follows in the scope of the claims. All publications, patents, and patent applications mentioned in the above specification are hereby incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference in its entirety.

Detailed descriptions of one or more preferred embodiments are provided herein. It is to be understood, however, that the present invention may be embodied in various forms. Therefore, specific details disclosed herein are not to be interpreted as limiting, but rather as a basis for the claims and as a representative basis for teaching one skilled in the art to employ the present invention in any appropriate manner. Indeed, various modifications of the described modes for carrying out the invention that are obvious to those skilled in the art are intended to be within the scope of the invention.

Other embodiments are in the claims. This application claims the benefit of U.S. Provisional Serial Nos. (i) 63/605,835, filed 04 December

2023, (ii) 63/606,057, filed 04 December 2023, (Hi) 63/569,327, filed 25 March 2024, and (iv) 63/711 ,529, filed 24 October 2024, each of which is incorporated in its entirety herein.

Claims

1 . A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV population having a surface cargo comprising an EV-associated form of a disease biomarker; b) using a first binding agent that preferentially binds to the EV-associated form of a disease biomarker cargo to measure the presence of, or level of, the EV population in the sample; c) measuring a total level of all EVs present in the mixture of EVs; and d) on the basis of steps (b) and (c), diagnosing the subject.

2. The method of claim 1 , wherein the total level of all EVs present in the mixture of EVs is measured using light scattering, atomic force microscopy, scanning electron microscopy, flow cytometry, surface plasmon resonance, biolayer interferometry, immunoassays, and/or lipid/protein staining.

3. The method of claims 1 or 2, wherein: i the EV-associated form of the disease biomarker comprises Ap, Ap 42, GFAP, and/or Tau; and ii the disease is an amyloidosis.

4 The method of claims 1 or 2, wherein: i the EV-associated form of the disease biomarker comprises GFAP, NfL, NSE, and/or SWOB; and ii the disease is a brain vascular damage.

5 The method of claims 1 or 2, wherein: i the EV-associated form of the disease biomarker comprises aSyn, paSyn129, GFAP, and/or NfL; and ii the disease is a synucleinopathy.

6 The method of claims 1 or 2, wherein: i the EV-associated form of the disease biomarker comprises Tau, pTau231 , and/or pTau396; and ii the disease is a tauopathy.

7 The method of claims 1 or 2, wherein: i the EV-associated form of the disease biomarker comprises aSyn, paSyn, GFAP, NfL, TDP43, and/or pTDP43; and ii the disease is a TDP43 proteinopathy.

8 A method of diagnosing disease in a subject in need thereof, the method comprising: a providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV population having a first surface cargo comprising an EV-associated form of a first marker cargo and a second EV population having a second surface cargo comprising an EV-associated form of a second marker cargo; b) using a first binding agent that preferentially binds to the EV-associated form of the first marker cargo to measure the level of the first EV population in the sample; c) using a second binding agent that preferentially binds to the EV-associated form of the second marker cargo to measure the level of the second EV population in the sample; and d) on the basis of steps (b) and (c), diagnosing the subject, wherein the first EV population and the second EV population are different.

9. The method of claim 8, wherein the second binding agent is a first pan binding agent and the level of the second EV population in the sample is a total level of all EVs present in the mixture of EVs, wherein the first pan binding agent specifically binds a pan marker.

10. The method of claims 8 or 9, wherein step c) further comprises using a third binding agent that preferentially binds to an EV-associated form of a third marker cargo to measure the level of the second EV population in the sample, wherein the third binding agent is a second pan binding agent, wherein the second binding agent specifically binds a pan marker.

11 . The method of claim 10, wherein the first pan binding agent and the second pan binding agent are different.

12. The method of any one of claims 9-11 , wherein the pan marker is CD9 or CD81 .

13. The method of claim any one of claims 8-12, wherein step d) further comprises calculating the value of step b) normalized to the value of step c).

14. The method of claim any one of claims 8-12, wherein step d) further comprises calculating the value of step c) normalized to the value of step b).

15. The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises Ap, Ap 42, GFAP, and/or Tau; and ii the disease is an amyloidosis.

16 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises GFAP, NfL, NSE, and/or SWOB; and ii the disease is a brain vascular damage.

17 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises aSyn, paSyn129, GFAP, and/or NfL; and ii the disease is a synucleinopathy.

18. The method of any one of claims 8-14, wherein: i) the EV-associated form of the first marker cargo comprises Tau, pTau231 , and/or pTau396; and ii) the disease is a tauopathy.

19. The method of any one of claims 8-14, wherein: i) the EV-associated form of the first marker cargo comprises TDP43 and/or pTDP43; and ii the disease is a TDP43 proteinopathy.

20 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and ii the disease is a brain vascular damage.

21 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and ii the disease is a synucleinopathy.

22 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and ii the disease is a tauopathy.

23 The method of any one of claims 8-14, wherein: i the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or SYP; and ii the disease is a TDP43 proteinopathy.

24 The method of claim 8, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; ii the EV-associated form of the second marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and iii the disease is a brain vascular damage.

25 The method of claim 7, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; ii the EV-associated form of the second marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and iii the disease is a synucleinopathy.

26 The method of claim 7, wherein: i the EV-associated form of the first marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; ii) the EV-associated form of the second marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and iii) the disease is a tauopathy.

27. The method of claim 7, wherein: i) the EV-associated form of the first marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; ii) the EV-associated form of the second marker cargo includes GLAST, NrCAM, CD171 , and/or CD56; and Hi) the disease is a TDP43 proteinopathy.

28. The method of any one of claims 24-27, wherein step c) further comprises using a third binding agent that preferentially binds to an EV-associated form of a third marker cargo to measure the level of the second EV population in the sample, wherein the third binding agent is a second pan binding agent, wherein the second pan binding agent specifically binds a pan marker.

29. The method of any one of claims 8-28, wherein step b) further comprises using a fourth binding agent that preferentially binds to an EV-associated form of a fourth marker cargo to measure the level of the first EV population in the sample, wherein the fourth binding agent is a third pan binding agent, wherein the fourth binding agent specifically binds a pan marker.

30. The method of claim 28 or 29, wherein the pan marker is CD9 or CD81 .

31 . A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) using a first binding agent that preferentially binds to the source marker cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) using a second binding agent that preferentially binds to a disease biomarker cargo on the EV subpopulation complexed according to the source marker cargo; d) measuring the presence of, or level of, the disease biomarker cargo on the EV subpopulation complexed according to the source marker cargo; and e) on the basis of step d), diagnosing the subject.

32. The method of claim 31 , wherein step c) further comprises permeabilizing the EV subpopulation complexed according to source marker with a mild nonionic surfactant to expose disease biomarker cargo on and in the EV subpopulation; and step d) further comprises measuring the level of disease biomarker cargo on and in the EV subpopulation complexed according to source marker.

33. The method of 32, wherein the surfactant is a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate.

34. The method of any one of claims 31 -33, wherein: x) the disease marker cargo comprises NSE and/or S100B; y) the source marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a brain vascular damage.

35. The method of any one of claims 31 -33, wherein: x) the disease marker cargo comprises aSyn, paSyn129, and/or GFAP; y) the source marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a synucleinopathy.

36. The method of any one of claims 31 -33, wherein: x) the disease marker cargo comprises Tau, pTau231 , and/or pTau396; y) the source marker cargo comprises GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a tauopathy.

37. The method of any one of claims 31 -33, wherein: x) the disease marker cargo comprises TDP43 and/or pTDP43; y) the source marker cargo comprises SYP, GLAST, NrCAM, CD171 , and/or CD56; and z) the disease is a TDP43 proteinopathy.

38. The method of any one of claims 31 -37, further comprising:

(f) using the first binding agent that preferentially binds to the source marker cargo on the EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker;

(g) using a third binding agent that preferentially binds to a pan marker cargo on the EV subpopulation complexed according to source marker;

(h) measuring the level of pan marker on the EV subpopulation complexed according to the source marker;

(i) on the basis of steps (d) and (h), diagnosing the subject.

39. The method of claim 38, wherein step (i) further comprises calculating the value of step (d) normalized to the value of step (h).

40. The method of claim 38, wherein step (i) further comprises calculating the value of step (h) normalized to the value of step (d).

41 . The method of claim 38, wherein the pan marker is CD81 or CD9.

42. The method of any one of claims 31 -33, further comprising:

(x) measuring the level of the disease biomarker NSE and/or S100B on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56,

(y) measuring the level of the disease biomarker NSE and/or S100B on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56,

(z) on the basis of steps (x) and (y), diagnosing the subject with a brain vascular damage, wherein the first EV subpopulation and the second EV subpopulation are different.

43. The method of any one of claims 31 -33, further comprising:

(x) measuring the level of the disease biomarker aSyn, paSyn129, and/or GFAP on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56;

(y) measuring the level of the disease biomarker aSyn, paSyn129, and/or GFAP on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56;

(z) on the basis of steps (x) and (y), diagnosing the subject with a synucleinopathy, wherein the first EV subpopulation and the second EV subpopulation are different.

44. The method of any one of claims 31 -33, further comprising:

(x) measuring the level of the disease biomarker Tau, pTau231 , and/or pTau396 on the first EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56;

(y) measuring the level of the disease biomarker Tau, pTau231 , and/or pTau396 on a second EV subpopulation complexed according to the source marker GLAST, NrCAM, CD171 , and/or CD56;

(z) on the basis of steps (x) and (y), diagnosing the subject with a tauopathy, wherein the first EV subpopulation and the second EV subpopulation are different.

45. The method of any one of claims 31 -33, further comprising:

(x) measuring the level of the disease biomarker TDP43 and/or pTDP43 on the first EV subpopulation complexed according to the source marker SYP, GLAST, NrCAM, CD171 , and/or CD56;

(y) measuring the level of the disease biomarker TDP43 and/or pTDP43 on a second EV subpopulation complexed according to the source marker SYP, GLAST, NrCAM, CD171 , and/or CD56;

(z) on the basis of steps (x) and (y), diagnosing the subject with ALS, wherein the first EV subpopulation and the second EV subpopulation are different.

46. The method of any one of claims 42-45, wherein step (z) further comprises calculating the value of step (x) normalized to the value of step (y).

47. A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject; b) measuring a level of a first EV population in the mixture of EVs; c) measuring a level of a second EV population in the mixture of EVs; d) calculating a first composite value of step (b) normalized to the value of step (c); e) measuring a level of a third EV population in the mixture of EVs; f) measuring a level of a fourth EV population in the mixture of EVs; g) calculating a second composite value of step (e) normalized to the value of step (f); h) comprises combining each composite value of steps (d) and (g) into an algorithm classifier for use in differentially diagnosing the subject; and i) on the basis of step (h) and the algorithm, diagnosing the subject with the disease, wherein the first EV population and the second EV population are different, and wherein the third EV population and the fourth EV population are different.

48. The method of claim 47, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 4; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 4; iii the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 4; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 4; and v the disease is an amyloidosis.

49 The method of claim 47, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 8 or 16; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 8 or 16 iii the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 8 or 16; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 8 or 16 and v the disease is a brain vascular damage.

50 The method of claim 47, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 12 or 21 ; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 12 or 21 iii the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 12 or 21 iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 12 or 21 and v the disease is a synucleinopathy.

51 The method of claim 47, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 26; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 26; iii the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 26; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 26; and v) the disease is a tauopathy.

52. The method of claim 47, wherein: i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 31 ; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 31 ;

Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 31 ; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 31 ; and v) the disease is ALS.

53. The method of any one of claims 47-52, wherein after step g), the method further comprises: g1 ) measuring a level of a fifth EV population in the mixture of EVs; g2) measuring a level of a sixth EV population in the mixture of EVs; and g3) calculating a third composite value of step g1 ) normalized to the value of step g2); and wherein step h) comprises combining each composite value of steps d), g), and g3) into an algorithm classifier for use in differentially diagnosing the subject, wherein the fifth EV population and the sixth EV population are different.

54. The method of claim 53, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 5; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 5;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 5; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 5; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 5; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 5; and vii the disease is an amyloidosis.

55 The method of claim 53, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 9 or 17; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 9 or 17

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 9 or 17; and iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 9 or 17 v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of any one of Tables 9 or 17; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of any one of Tables 9 or 17; and vii the disease is a brain vascular damage.

56 The method of claim 53, wherein: i) the first EV subpopulation is any EV subpopulation listed under Markers 1 A of any one of Tables 13 or 23; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of any one of Tables 13 or 23;

Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of any one of Tables 13 or 23; and iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of any one of Tables 13 or 23; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of any one of Tables 13 or 23; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of any one of Tables 13 or 23; and vii) the disease is a synucleinopathy.

57. The method of claim 53, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 27; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 27;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 27; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 27; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 27; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 27; and vii the disease is a tauopathy.

58 The method of claim 53, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 32; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 32;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 32; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 32; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 32; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 32; and vii the disease is ALS.

59 The method of any one of claims 53-58, wherein after step g3), the method further comprises: g4 measuring a level of a seventh EV population in the mixture of EVs; g5 measuring a level of an eighth EV population in the mixture of EVs; and g6 calculating a third composite value of step g4) normalized to the value of step g5); and wherein step h) comprises combining each composite value of steps d), g), g3), and g6) into an algorithm classifier for use in differentially diagnosing the subject, wherein the seventh EV population and the eighth EV population are different.

60 The method of claim 59, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1A of Table 18; ii) the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 18;

Hi) the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 18; iv) the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 18; v) the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 18; vi) the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 18; vii) the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 18; viii) the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 18; and ix) the disease is a brain vascular damage.

61 . The method of claim 59, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 23; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 23;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 23; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 23; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 23; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 23; vii the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 23; viii the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 23; and ix the disease is a synucleinopathy.

62 The method of claim 59, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 28; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 28;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 28; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 28; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 28; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 28; vii the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 28; viii the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 28; and ix the disease is a tauopathy.

63 The method of claim 59, wherein: i the first EV subpopulation is any EV subpopulation listed under Markers 1 A of Table 33; ii the second EV subpopulation is any EV subpopulation listed under Markers 1 B of Table 33;

Hi the third EV subpopulation is any EV subpopulation listed under Markers 2A of Table 33; iv the fourth EV subpopulation is any EV subpopulation listed under Markers 2B of Table 33; v the fifth EV subpopulation is any EV subpopulation listed under Markers 3A of Table 33; vi the sixth EV subpopulation is any EV subpopulation listed under Markers 3B of Table 33; vii the seventh EV subpopulation is any EV subpopulation listed under Markers 44A of Table 33; viii the eighth EV subpopulation is any EV subpopulation listed under Markers 4B of Table 33; and ix) the disease is ALS.

64. The method of any one of claims 47-63, wherein the algorithm comprises a non-linear model, wherein the non-linear model comprises a multiple logistic regression classifier, a support vector machine, and/or a random forest.

65. The method of claim 64, wherein the non-linear model further comprises feature selection algorithms, wherein the feature selection algorithms comprise a forward selection, a recursive feature elimination, and/or a penalized regression algorithm.

66. The method of any one of claims 1 -65, wherein the sample is substantially free of exogenous EVs.

67. The method of any one of claims 1 -66, wherein prior to step (b), the method further comprises diluting the sample with a diluent, wherein the diluent comprises a protein.

68. The method of claim 67, wherein the protein comprises IgG, IgA, and/or IgM to competitively inhibit nonspecific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent.

69. The method of claim 67 or 68, wherein the diluent further comprises a polymer to increase the preferential binding in steps (b) and (c) by altering the viscosity of the sample and/or inducing the macromolecular crowding effect in the sample, wherein the polymer comprises polyethylene glycol, polyvinylpyrrolidone, dextran, mannitol, betaine, mannitol, sorbitol, xylitol, or other commonly known and used stabilizers.

70. The method of any one of claims 67-69, wherein the diluent further comprises a preservative to maintain a long-term sterility of the sample, wherein the preservative comprises any of sodium azide, ProCiin™. Thimerosal, Sodium Benzoate, or other commonly known and used preservatives.

71 . The method of any one of claims 67-70, wherein the diluent further comprises a detergent to substantially reduce non-specific binding to a surface.

72. The method of any one of claims 1 -71 , wherein the method further comprises diluting the sample with a diluent, wherein the diluent is substantially free of exogenous EVs.

73. A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) complexing a first binding agent that preferentially binds to the source marker cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) permeabilizing the EV subpopulation complexed according to the source marker with a mild nonionic surfactant to expose the disease biomarker cargo on the surface of and/or within the EV subpopulation complexed according to the source marker cargo; d) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the source marker cargo; e) measuring the presence of, or level of, the disease biomarker cargo in the permeabilized EV subpopulation complexed according to the source marker cargo; and f on the basis of step e), diagnosing the subject.

74 A method of diagnosing disease in a subject in need thereof, the method comprising: a providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a first cargo, and (ii an EV-associated form of a second cargo; b complexing a first binding agent that preferentially binds to the first cargo on the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the first cargo; c permeabilizing the EV subpopulation complexed according to the first cargo with a mild nonionic surfactant to expose the second cargo on the surface of and/or within the EV subpopulation complexed according to the first cargo; d complexing a second binding agent that preferentially binds to the second cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the first cargo; e measuring the presence of, or level of, the second cargo in the permeabilized EV subpopulation complexed according to the first cargo; and f on the basis of step e), diagnosing the subject.

75 A method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method comprising: a providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface comprising the EV-associated biomarker cargo; b diluting the sample with a diluent comprising a protein, wherein the diluent is substantially free of exogenous EVs; c following step (b), capturing the EV subpopulation having the surface comprising the EV-associated biomarker cargo; d washing the surface with a mild nonionic surfactant to permeabilize the captured EV subpopulation; e using a binding agent that preferentially binds to the EV-associated biomarker cargo on the surface of and/or within the captured and permeabilized EV subpopulation; and f measuring the presence of, or level of, the EV-associated biomarker cargo.

76 A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) diluting the sample with a diluent comprising a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further comprises a mild nonionic surfactant to permeabilize the first EV subpopulation; c) complexing a first binding agent that preferentially binds to the source marker cargo on the surface of and/or within the permeabilized first EV subpopulation from the mixture of EVs in the sample to form a permeabilized EV subpopulation complexed according to the source marker cargo; d) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the source marker cargo; e) measuring the presence of, or level of, the disease biomarker cargo in the permeabilized EV subpopulation complexed according to the source marker cargo; and f on the basis of step e), diagnosing the subject.

77 A method of diagnosing disease in a subject in need thereof, the method comprising: a providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a first cargo, and (ii an EV-associated form of a second cargo; b diluting the sample with a diluent comprising a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further comprises a mild nonionic surfactant to permeabilize the first EV subpopulation; c complexing a first binding agent that preferentially binds to the first cargo on the surface of and/or within the permeabilized first EV subpopulation from the mixture of EVs in the sample to form a permeabilized EV subpopulation complexed according to the first cargo; d complexing a second binding agent that preferentially binds to the second cargo on the surface of and/or within the permeabilized EV subpopulation complexed according to the first cargo; e measuring the presence of, or level of, the second cargo in the permeabilized EV subpopulation complexed according to the first cargo; and f on the basis of step e), diagnosing the subject.

78 A method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method comprising: a providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface comprising the EV-associated biomarker cargo; b diluting the sample with a diluent comprising a protein, wherein the diluent is substantially free of exogenous EVs, wherein the diluent further comprises a mild nonionic surfactant to permeabilize the EV subpopulation; c following step (b), capturing the permeabilized EV subpopulation having the EV-associated biomarker cargo on the surface of and/or within the permeabilized EV subpopulation; d) using a binding agent that preferentially binds to the EV-associated biomarker cargo on the surface of and within the captured and permeabilized EV subpopulation; and e) measuring the presence of, or level of, the EV-associated biomarker cargo.

79. The method of any one of claims 73-78, wherein the mild ionic surfactant is a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate.

80. The method of any one of claims 73-79, wherein the mild ionic surfactant is a polysorbate surfactant selected from polyoxyethylene 20 sorbitan monolaurate, polyoxyethylene (4) sorbitan monolaurate, polyoxyethylene 20 sorbitan monopalmitate, polyoxyethylene 20 sorbitan monostearate; and polyoxyethylene 20 sorbitan monooleate, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.01 to 0.75% (w/w) polysorbate surfactant.

81 . The method of any one of claims 73-79, wherein the mild ionic surfactant is a polyethylene glycol alkyl ether selected from PEG-2 oleyl ether, oleth-2; PEG-3 oleyl ether, oleth-3; PEG-5 oleyl ether, oleth-5; PEG- 10 oleyl ether, oleth-10; PEG-20 oleyl ether, oleth-20; PEG-4 lauryl ether, laureth-4; PEG-9 lauryl ether; PEG- 23 lauryl ether, laureth-23; PEG-2 cetyl ether; PEG-10 cetyl ether; PEG-20 cetyl ether; PEG-2 stearyl ether; PEG-10 stearyl ether; Polyoxyethylene (20) oleyl ether; PEG-20 stearyl ether; and PEG-100 stearyl ether, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.01 to 2.5% (w/w) polyethylene glycol alkyl ether.

82. The method of any one of claims 73-79, wherein the mild ionic surfactant is an alkylphenol ethoxylate selected from polyethylene glycol tert-octylphenyl ether and 2-[4-(2,4,4-trimethylpentan-2- yl)phenoxy]ethanol, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.1 to 2.5% (w/w) alkylphenol ethoxylate.

83. The method of any one of claims 75-78, wherein the protein comprises IgG, IgA, and/or IgM to competitively inhibit non-specific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent.

84. A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a disease biomarker cargo, and (ii) an EV-associated form of a source marker cargo characteristic of a cell type of origin; b) complexing a first binding agent that preferentially binds to the source marker cargo on the surface of and/or within the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the source marker cargo; c) complexing a second binding agent that preferentially binds to the disease biomarker cargo on the surface of the EV subpopulation complexed according to the source marker cargo; d) measuring the presence of, or level of, the disease biomarker cargo in the EV subpopulation complexed according to the source marker cargo; and e) on the basis of step d), diagnosing the subject.

85. A method of diagnosing disease in a subject in need thereof, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes a first EV subpopulation having a surface comprising (i) an EV-associated form of a first cargo, and (ii) an EV-associated form of a second cargo; b) complexing a first binding agent that preferentially binds to the first cargo on the surface of the first EV subpopulation from the mixture of EVs in the sample to form an EV subpopulation complexed according to the first cargo; c) complexing a second binding agent that preferentially binds to the second cargo on the surface of the EV subpopulation complexed according to the first cargo; d) measuring the presence of, or level of, the second cargo in the EV subpopulation complexed according to the first cargo; and e) on the basis of step d), diagnosing the subject.

86. A method of detecting an EV-associated biomarker cargo in a mixture of EVs, the method comprising: a) providing a sample comprising a mixture of EVs obtained from the subject, wherein the mixture of EVs includes an EV subpopulation having a surface comprising the EV-associated biomarker cargo; b) following step (a), capturing the EV subpopulation having the EV-associated biomarker cargo on the surface of the EV subpopulation; c) using a binding agent that preferentially binds to the EV-associated biomarker cargo on the surface the captured EV subpopulation; and d) measuring the presence of, or level of, the EV-associated biomarker cargo.

87. The method of any one of claims 84-86, further comprising adding a diluent comprising a mild nonionic surfactant to permeabilize the EV subpopulation, wherein the step of adding the diluent follows (i) step a); (ii) step b), and/or (iii) step c).

88. The method of claim 87, wherein the mild ionic surfactant is a polysorbate surfactant, a polyethylene glycol alkyl ether, or an alkylphenol ethoxylate.

89. The method of claim 87, wherein the mild ionic surfactant is a polysorbate surfactant selected from polyoxyethylene 20 sorbitan monolaurate, polyoxyethylene (4) sorbitan monolaurate, polyoxyethylene 20 sorbitan monopalmitate, polyoxyethylene 20 sorbitan monostearate; and polyoxyethylene 20 sorbitan monooleate, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.01 to 0.75% (w/w) polysorbate surfactant.

90. The method of claim 87, wherein the mild ionic surfactant is a polyethylene glycol alkyl ether selected from PEG-2 oleyl ether, oleth-2; PEG-3 oleyl ether, oleth-3; PEG-5 oleyl ether, oleth-5; PEG-10 oleyl ether, oleth-10; PEG-20 oleyl ether, oleth-20; PEG-4 lauryl ether, laureth-4; PEG-9 lauryl ether; PEG-23 lauryl ether, laureth-23; PEG-2 cetyl ether; PEG-10 cetyl ether; PEG-20 cetyl ether; PEG-2 stearyl ether; PEG-10 stearyl ether; Polyoxyethylene (20) oleyl ether; PEG-20 stearyl ether; and PEG-100 stearyl ether, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.01 to 2.5% (w/w) polyethylene glycol alkyl ether.

91 . The method of claim 87, wherein the mild ionic surfactant is an alkylphenol ethoxylate selected from polyethylene glycol tert-octylphenyl ether and 2-[4-(2,4,4-trimethylpentan-2-yl)phenoxy]ethanol, and wherein the permeabilizing step comprises exposing the EV to a solution comprising from about 0.1 to 2.5% (w/w) alkylphenol ethoxylate.

92. The method of claim 87, wherein the diluent further comprises a protein, wherein the protein comprises IgG, IgA, and/or IgM to competitively inhibit non-specific binding of one or more biomolecules in the sample to the first binding agent and/or the second binding agent.

93. The method of any one of claims 1 -92, wherein the step of measuring the presence of, or level of, the EV population and/or biomarker in the sample comprises measuring a detectable signal that is a fluorescent, chemiluminescent, radiological, or colorimetric signal.

94. The method of claim 93, wherein the step of measuring involves a direct ELISA, an indirect ELISA, a sandwich ELISA, or a competitive ELISA-based assay.