WO2025024790A1 - Digital assay for single extracellular vesicle, single molecule detection - Google Patents

Abstract

A method, comprising: releasing at least one protein carried by an extracellular vesicle (EV) while the EV is in a container free of other EVs; and collecting a signal associated with the at least one protein. A method, comprising: with an EV immobilized in a matrix, collecting a signal indicative of a labeled probe molecule being bound to a pendant nucleic acid associated with the EV. A method, comprising: immobilizing an EV within a matrix; contacting the EV with an antibody that binds to a protein associated with the EV, wherein the antibody is conjugated to a pendant nucleic acid; amplifying the pendant nucleic acid; contacting the pendant nucleic acid with a probe comprising a fluorophore conjugated to a probe nucleic acid, the probe nucleic acid being complementary to the pendant nucleic acid; collecting a signal indicative of the probe being bound to the protein.

Description

DIGITAL ASSAY FOR SINGLE EXTRACELLULAR

VESICLE, SINGLE MOLECULE DETECTION

RELATED APPLICATIONS

[0001] The present application claims priority to and the benefit of United States patent application no. 63/515,682, “Double Digital Assay For Single Extracellular Vesicle, Single Molecule Detection” (filed July 26, 2023). All foregoing applications are incorporated herein by reference in their entireties for any and all purposes.

GOVERNMENT RIGHTS

[0002] This invention was made with government support under CA256353 awarded by the National Institutes of Health. The government has certain rights in the invention.

TECHNICAL FIELD

[0003] The present disclosure relates to the field of extracellular vesicles and to the field of molecular analysis systems.

BACKGROUND

[0004] Extracellular vesicles (EVs) have emerged as a promising source of biomarkers for disease diagnosis. However, current diagnostic methods for EVs present formidable challenges, given the low expression levels of biomarkers carried by EV samples, as well as their complex physical and biological properties. Accordingly, there is a need in the art for improved methods of EV detection as well as methods of detecting biomarkers carried by EVs.

SUMMARY

[0005] Herein, a highly sensitive double digital assay that allows for the absolute quantification of individual molecules from a single EV was developed. Because the relative abundance of proteins is low for a single EV, tyramide signal amplification (TSA) was integrated to increase the fluorescent signal readout for evaluation. With the integrative microfluidic technology, the technology’s ability to compartmentalize single EVs and digital partitioning capacity were successfully demonstrated. The device was then applied to detect single PD-L1 proteins from single EVs derived from a melanoma cell line and discovered that there are ~2.7 molecules expressed per EV, demonstrating the applicability of the system for profiling important prognostic and diagnostic cancer biomarkers for therapy response, metastatic status, and tumor progression. The ability to accurately quantify protein molecules of rare abundance from individual EVs will shed light on the understanding of EV heterogeneity and discovery of EV subtypes as new biomarkers.

[0006] In one aspect, the disclosed technology provides a method, comprising: releasing at least one protein carried by an extracellular vesicle (EV) while the EV is in a container free of other EVs; and collecting a signal associated with the at least one protein.

[0007] Also provided is a method, comprising: partitioning a plurality of EVs among a plurality of containers such that a given partitioned EV is in a container free of other EVs; releasing proteins from at least some of the partitioned EVs; collecting signals from the released proteins.

[0008] Further provided is a system, comprising: a plurality of containers, the containers optionally comprised in a substrate; and a container having disposed therein (i) capture microbeads configured to bind to at least one protein released by an EV and (ii) spacer microbeads configured to be free of binding to the at least one protein, and wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

[0009] Further provided is a method, comprising: with an EV immobilized in a matrix, collecting a signal indicative of a labeled probe molecule being bound to a pendant nucleic acid associated with the EV.

[0010] Additionally disclosed is a method, comprising: immobilizing an EV within a matrix; contacting the EV with an antibody that binds to a protein associated with the EV, wherein the antibody is conjugated to a pendant nucleic acid; amplifying the pendant nucleic acid; contacting the pendant nucleic acid with a probe comprising a fluorophore conjugated to a probe nucleic acid, the probe nucleic acid being complementary to the pendant nucleic acid; collecting a signal indicative of the probe being bound to the protein.

[0011] Also provided is a system, comprising: an EV immobilized within a matrix, a fluorescence imager, and a compression element, wherein the fluorescence imager collects a signal indicative of a probe associated with a protein, the protein being associated with the EV, and wherein the compression element is optionally configured to compress the matrix to a height of less than about 6 pm, less than about 8 pm, less than about 10 pm, less than about 12 pm, or less than about 14 pm.

[0012] Further provided is a system, comprising: an EV immobilized within a matrix, an antibody associated with the EV, the antibody being conjugated to a pendant nucleic acid; a probe molecule bound to the pendant nucleic acid, the probe molecule comprising a probe nucleic acid and a fluorophore.

BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In the drawings, which are not necessarily drawn to scale, like numerals may describe similar components in different views. Like numerals having different letter suffixes may represent different instances of similar components. The drawings illustrate generally, by way of example, but not by way of limitation, various aspects discussed in the present document. In the drawings:

[0014] FIG. 1. Device and schematic. A) Schematic of double digital EV microwell protein detection. B) CAD files of the bottom microwell layer and top flow layer. C) PDMS microwell devices: width (W) = 42 pm and distance (D) = 20 pm. (Scale bar 50 pm) D) Loading bonded microwell device with a solution in the reservoir and washing the system with a pump via connected tubing. E) Visualization of spacer streptavidin beads (90%) and capture streptavidin beads with conjugated biotin-NHS-Alexa Fluor (AF) 647 linker (10%) loaded into the microwell device. (Scale bar: 50 pm)

[0015] FIG. 2. TSA off-chip. A) EGFR protein capture and amplification from cell & EV lysis on epoxy beads: TSA, positive control (no TSA), and negative control (ctrl). (Scale bar = 20 pm) B) Quantification of mean fluorescence intensity (MFI = a.u.) from off- chip TSA.

[0016] FIG. 3. EV loading optimization for microwell array. A) Single-cell loading into microwells at different concentrations. Expected vs actual single-cell loading into individual microwells quantified. (Scale bar = 50 pm) B) A431 EV titration into microwell array with EGFR protein capture and detection. (Scale bar = 100 pm) C) Magnified images A431 EV titration into microwell array with fluorescent EGFR protein captured beads and spacer beads. (Scale bar = 12.5 pm) [0017] FIG. 4. PD-L1 protein detection in microwell array. A) Off-chip PD-L1 protein detection from PD-L1(+/-) EVs. Quantification of mean fluorescence intensity (MFI = a.u.) from off-chip TSA. (Scale bar = 20 pm) B) On-chip PD-L1 protein detection from PD-L1(+/-) EVs. Quantification of wells positive for signal coming from single EVs. (Scale bar = 12.5 pm)

[0018] FIG. 5. Digital counting of PD-L1 molecules per EV. A) Standard curve created using recombinant human PD-L1 proteins. Bulk PD-L1 (+/-) EVs loaded at different concentrations (10 pg/mL & 20 pg/mL) and detected with ELISA. B) Average number of PD-L1 molecules coming from single PD-L1(+/-) EVs based on ELISA. C) The number of positive beads per single PD-L1(+) EV (n=70). D) The average number of PD-L1 molecules per single PD-L1(+/-) EV. E) The percent positive beads for synthetic PD-L1 protein capture and detection. LOD reported for TSA (red) and ELISA (blue).

[0019] FIG. 6. Schematic for EV microwell assay.

[0020] FIG. 7. EV lysing with different conditions with/out sonication and Triton

X-100 (1% and 10%).

[0021] FIG. 8. Single-cell and EV loading into microwells. Lambda (X) is reported on each graph.

[0022] FIG. 9. PD-L1 protein detection off-chip with TSA, positive control (no

TSA), and negative control (ctrl). (Scale bar = 20 pm).

[0023] FIG. 10. PD-L1 (+/-) EV fluorescence imaging. A) Super-resolution imaging of PD-L1 (+/-) EVs. The number of blinking events was counted from doublepositive EVs (error bar = standard error). B) Fluorescence-based imaging of PD-L1 (+/-) EVs with an inverted fluorescence microscope.

[0024] FIG. 11. Overview of the multiplexed single EV analysis using MeHA hydrogel microparticles (MHPs). (a) Droplets including EVs and precursor, composed of MeHA and PI, are fabricated by microfluidic technique, (b) MHPs were prepared by exposing UV to droplets, which physically capture EVs. The synthesized MHPs were incubated with antibody -DNA (Ab-DNA) followed by rolling circle amplification (RCA) to amplify protein markers from EVs. (c) MHPs were squeezed to align the RCA dots on the same plane for the imaging, (d) Customized image analysis software automatically analyzes the presence or absence and colocalization of protein markers from single EVs. [0025] FIG. 12. Preparation and characterization of MHPs. (a) MHPs were produced by photocuring droplets in the polymerization chamber, (b) The UV exposure time was varied to evaluate the encapsulation efficiency of MHPs using 100 nm fluorescent beads as model EVs. (c) Assessment of remaining fluorescent beads in MHPs before (To) and after (Ti) the assay procedure according to the different UV exposure times. Error bars indicate the standard deviation from five MHPs (d) Force-indentation curves and (e) Young’s modulus for hydrogels synthesized by PEG and two different concentrations of MeHA. (f) Micrographs and (g) diameter of MHPs before and after squeezing. Error bars indicate the standard deviation from ten particles.

[0026] FIG. 13. RCA-based assay validation and optimization, (a) Schematic of the RCA assay within the hydrogel networks, (b) Optimization of Ab-DNA concentration. The signal -to-noise ratio (SNR) was estimated as the ratio of RCA dots between targeted and control (no targeted protein) MHPs. Data are displayed as mean ±sd from five replicates, (c) Optimization of RCA time. The fluorescent intensity of MHPs (purple) and the number of RCA dots (black) were measured during amplification. Error bars indicate the standard deviation from five MHPs. (d) Fluorescence micrographs of targeted and control MHPs after RCA-based assay, (e) Assay validation. RCA dots were observed only when all assay components were included. Bars indicate the number of dots from five MHPs.

[0027] FIG. 14. Characterization of EVs using MHPs. (a) Calibration curve with varying amounts of A431 EVs. Error bars are standard deviations from seven MHPs. (b, c) RCA probe sets specificity. Fluorescence micrographs and bar graphs demonstrated high specificity of the designed probe sets. Error bars are standard deviations from seven MHPs. (d) Evaluation of assay specificity. Bars indicate the number of RCA dots from A431 EVs from ten MHPs. (e) Multiplexed analysis of A431 EVs. Venn diagram showed percentage of single, double, and triple positive EVs among the tested markers, (f) Fluorescence micrographs of MHPs after 3-plex assay, (g) Mapping of single EV. Each row represents a single biomarker, and each column represents a single EV. (h) Comparison of co-expression ratio from solution, hydrogel, and surface assay by analyzing CD9 and EGFR from A431 EVs.

[0028] FIG. 15. Multiplexed single EV profiling reveals heterogeneity of pancancer marker expression among EVs shed by different cancer cell lines, (a) Molecular protein expression of pan-cancer markers (EGFR, EpCAM, and MUC1) and EV tetraspanin marker CD9 was characterized on parent cells and EVs derived from 4 cancer cell lines (A431, A549, MCF7, and PANCI), (b) Correlation of cancer protein marker expression between EVs and respective parental cell lines, (c) Analysis of biomarker concurrence (right; legend of protein marker combinations) across single EVs obtained from all four cell lines.

[0029] FIG. 16. 1H NMR spectra of MeHA. Spectra were acquired at 500 MHz at a concentration of 5 mg/mL in D2O. The degree of substitution (30%) was determined by integration of the vinyl group (5=6.05, 1 H and 5=5.66, 1 H) relative to the HA backbone (5=3.0-4.0, 10 H).

[0030] FIG. 17. Generation of droplets, (a) A micrograph of a flow-focusing microfluidic droplet generator, (b) A merged micrograph of the inlet area. Fluorophore conjugated acrydite-DNAs were added on either side and MeHA was added in the center, mimicking the droplet generation process, (c) Fluorescence micrograph and (d) fluorescence plot of the hydrogels confirms the uniform mixing of the precursor components.

[0031] FIG. 18. Before and after squeezing of PHPs. There was no change in diameter after squeezing PHPs. Error bars indicate the standard deviation from ten particles.

[0032] FIG. 19. Characterization and utilization of remaining unreacted methacrylate groups in MHPs. (a) Schematic of the reaction between unreacted methacrylate groups and FAM-PEG-SH. Through thiol-ene Michael addition reaction, thiol groups are bind to electro-deficient unreacted methacrylate groups, resulting in fluorescent labeling of methacrylate groups, (b) Validation of the remaining unreacted methacrylate groups in the hydrogel. To prove that the fluorescent labeling is due to the thiol-ene reaction, one group of MHPs was treated with PEG-SH to block the unreacted methacrylate groups prior to reaction with FAM-PEG-SH. MHPs without blocking showed higher fluorescence whereas blocked MHPs showed no fluorescence. Error bars indicate the standard deviation from five MHPs. (c-d) Validation of the reaction between protein molecules and unreacted methacrylate groups. Primary amine groups of protein molecules from cell lysates form C-N bond with methacrylate groups by aza-Michael addition reaction. After reaction, RCA was performed by targeting EGFR molecules. Non-blocked MHPs showed amplified dots whereas blocked MHPs showed no dots. Error bars indicate the standard deviation from five MHPs.

[0033] FIG. 20. Analysis of RCA dots within the MHPs. Customized software identified the RCA dots from the fluorescence images taken from different fluorescence channels and then generated contours for counting the RCA dots. Identified contours from each fluorescence channel were combined into a single image to find the colocalization between each dot.

[0034] FIG. 21. Cyclic imaging of MeHA hydrogel particles (MHPs). (a) CD9 and EGFR were targeted and amplified by RCA in A431 EV-loaded MHPs. After amplification, FAM-DNA targeting CD9 amplicons were hybridized as a first round. Then, fluorophores were stripped off by DNA denaturation using 50% formamide for 10 min. In the second round, Cy5-DNA targeting EGFR amplicons were hybridized, (b) Validation of the stability of cyclic imaging. Bars indicate the number of EGFR RCA dots with and without cyclic imaging. Error bars indicate the standard deviation from seven MHPs.

DETAILED DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS

[0035] The present disclosure may be understood more readily by reference to the following detailed description of desired embodiments and the examples included therein.

[0036] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art. In case of conflict, the present document, including definitions, will control. Preferred methods and materials are described below, although methods and materials similar or equivalent to those described herein can be used in practice or testing. All publications, patent applications, patents and other references mentioned herein are incorporated by reference in their entirety. The materials, methods, and examples disclosed herein are illustrative only and not intended to be limiting.

[0037] The singular forms “a,” “an,” and “the” include plural referents unless the context clearly dictates otherwise.

[0038] As used in the specification and in the claims, the term "comprising" can include the embodiments "consisting of' and "consisting essentially of.” The terms “comprise(s),” “include(s),” “having,” “has,” “can,” “contain(s),” and variants thereof, as used herein, are intended to be open-ended transitional phrases, terms, or words that require the presence of the named ingredients/steps and permit the presence of other ingredients/steps. However, such description should be construed as also describing compositions or processes as "consisting of' and "consisting essentially of' the enumerated ingredients/steps, which allows the presence of only the named ingredients/steps, along with any impurities that might result therefrom, and excludes other ingredients/steps. [0039] As used herein, the terms “about” and “at or about” mean that the amount or value in question can be the value designated some other value approximately or about the same. It is generally understood, as used herein, that it is the nominal value indicated ±10% variation unless otherwise indicated or inferred. The term is intended to convey that similar values promote equivalent results or effects recited in the claims. That is, it is understood that amounts, sizes, formulations, parameters, and other quantities and characteristics are not and need not be exact, but can be approximate and/or larger or smaller, as desired, reflecting tolerances, conversion factors, rounding off, measurement error and the like, and other factors known to those of skill in the art. In general, an amount, size, formulation, parameter or other quantity or characteristic is “about” or “approximate” whether or not expressly stated to be such. It is understood that where “about” is used before a quantitative value, the parameter also includes the specific quantitative value itself, unless specifically stated otherwise.

[0040] Unless indicated to the contrary, the numerical values should be understood to include numerical values which are the same when reduced to the same number of significant figures and numerical values which differ from the stated value by less than the experimental error of conventional measurement technique of the type described in the present application to determine the value.

[0041] All ranges disclosed herein are inclusive of the recited endpoint and independently of the endpoints. The endpoints of the ranges and any values disclosed herein are not limited to the precise range or value; they are sufficiently imprecise to include values approximating these ranges and/or values.

[0042] As used herein, approximating language can be applied to modify any quantitative representation that can vary without resulting in a change in the basic function to which it is related. Accordingly, a value modified by a term or terms, such as “about” and “substantially,” may not be limited to the precise value specified, in some cases. In at least some instances, the approximating language can correspond to the precision of an instrument for measuring the value. The modifier “about” should also be considered as disclosing the range defined by the absolute values of the two endpoints. For example, the expression “from about 2 to about 4” also discloses the range “from 2 to 4.” The term “about” can refer to plus or minus 10% of the indicated number. For example, “about 10%” can indicate a range of 9% to 11%, and “about 1” can mean from 0.9-1.1. Other meanings of “about” can be apparent from the context, such as rounding off, so, for example “about 1” can also mean from 0.5 to 1.4. Further, the term “comprising” should be understood as having its open- ended meaning of “including,” but the term also includes the closed meaning of the term “consisting.” For example, a composition that comprises components A and B can be a composition that includes A, B, and other components, but can also be a composition made of A and B only. Any documents cited herein are incorporated by reference in their entireties for any and all purposes.

[0043] Any embodiment or aspect provided herein is illustrative only and does not limit the scope of the present disclosure or the appended claims. Any part or parts of any one or more embodiments of aspects can be combined with any part or parts of any one or more other embodiments or aspects.

[0044] Extracellular vesicles (EVs) are a group of heterogeneous lipid-bound nanoparticles (30 - 200 nm exosomes, <1000 nm microvesicles, and >1000 nm apoptotic bodies) that are actively shed by cells in both healthy and pathological states.^1,2 Because EVs are involved in intracellular communication and exhibit high stability for protecting their molecular cargo (DNA, RNA, and proteins), they have emerged as promising diagnostic biomarkers for different types of cancers, infectious diseases, and neurological disorders.¹'⁷ However, the diagnostic application of EVs is challenged by its population heterogeneity and lack of sensitive detection methods. Therefore, new technologies and strategies to improve EV-based diagnostics are warranted.

[0045] In recent years, several advanced methods have been developed to resolve EV heterogeneity by detecting and characterizing individual extracellular vesicles (EVs). These technologies include single EV analysis (sEVA) using microscopic imaging, modified flow cytometry for EV analysis, digital detection assays utilizing immunoaffinity capture (digital enzyme-linked immunosorbent assay; dELISA) of EVs, and nucleic acid-based amplification. Although these technologies have demonstrated remarkable success in profiling individual EVs, they have been capable of identifying the presence or absence of EV molecular cargo rather than providing absolute quantification of EV molecules. In order to accurately parse out EV heterogeneity and discover different EV subtypes, it is necessary not only to profile single EVs, but also to accurately quantify their molecules. However, single EV analysis has already been challenging due to the extremely low molecular content at the single EV level, and it becomes even more complicated when attempting to count individual molecules from single EVs. [0046] Digital assays are cutting-edge techniques that allow for the precise counting of single target biomolecules (e.g., proteins, nucleic acids) or entities such as EVs or cells. These systems rely on the Poisson distribution (X = 0.1) to prevent multiplets and digitally count analyte signals as single positive or negative events, achieving ultra-sensitive detection. While past studies have utilized multiple digital -based technologies to detect and quantify the number of molecules of disease biomarkers in clinical samples, no current method exists that can combine EV digital assays with single-molecule digital detection to achieve accurate biomarker expression levels at the single-EV and single-molecule resolution.

[0047] Herein, this paper presents a droplet-free double digital assay that utilizes bead-based microwell arrays and tyramide signal amplification (TSA) to achieve single EV, single molecule detection. It should be understood that although TSA was used to detect molecules carried by the EVs, the disclosed technology is not limited to the use of TSA. The technique involves the use of microwells to compartmentalize individual EVs and form a monolayer of microbeads for single-molecule capture and detection. By counting the number of fluorescent beads within the microwells, this platform can precisely profile the expression level of key EV biomarkers, improving our understanding of the composition of heterogeneous EV populations.

[0048] The working principle of the technology was shown by demonstrating the microwell array’s ability to partition a single EV per well. This was proven by loading and lysing A431 cell-derived EVs in the device and characterizing their epidermal growth factor receptor (EGFR) protein with TSA. After single EV loading validation was confirmed, we subsequently used our method to assess the abundance of the programmed death-ligand 1 (PD-L1) protein in a melanoma cell line (624-mel), proving the system's utility in detecting significant prognostic and diagnostic cancer markers. Together, we believe the presented double digital technology provides a new way of EV biomarker characterization and discovery by enabling absolute quantification of individual molecules from each EV.

[0049] Example Results and Discussion

[0050] Bead-based Digital Microwell Assays

[0051] Digital microwell assays have emerged as powerful tools for compartmentalizing and profiling single cells. In particular, these devices have enabled the genomic, transcriptomic, and proteinic profiling of these single cells, leading to new discoveries in biology and drug discovery. Microwells can be fabricated on a microfluidic device where they can be easily adapted to accommodate different cell types. Although their properties can be easily tuned, they have yet to be applied for single EV profiling.

[0052] To address this challenge, we have developed a highly sensitive digital assay that allows for the partitioning of single EVs and quantification of their protein abundance. The general workflow for our technology begins with bead loading into microwells. Here, microbeads were used to capture individual protein molecules from single EVs, achieving digital ELISA. To resolve individual bead signals, a mixture of two types of beads were prepared where there are beads coated with capture-targeting antibodies and other beads without antibody coating that serve as spacer beads. After bead loading, EV samples are loaded into the microwells and lysis buffer is flowed into the device, immediately followed by oil to prevent cross-contamination. After lysis, the TSA workflow is implemented for signal amplification. TSA is a technique for amplifying a signal that is weak or difficult to detect, allowing for the localization of the target molecule to a specific area. By depositing signal only at the site of amplification, TSA provides a localized signal that is contained to individual beads that can be easily recorded. (FIG. 1 A). An expanded workflow is provided in FIG. 6.

[0053] While some do not report many significant differences between these different shapes, honeycomb-based designs remain popular for their handling, high seeding efficacy (high surface area-to-volume ratio), and uniform spacing between wells. Therefore, we chose a honeycomb-inspired design to achieve optimal bead loading with minimal bead loss. Our device is divided into two layers: bottom layer (microwells) and top layer (flow chamber). (FIG. IB) The bottom layer contains a lattice of 10,000 microwells and the top flow chamber contains a pillar array to prevent itself from collapsing onto the microwells. The two layers were fabricated with photo- and soft-lithography to produce poly dimethylsiloxane (PDMS) devices. (FIG. 1C)

[0054] After PDMS casting and curing, the cut and hole-punched devices were then bonded together via plasma bonding to enclose the microwells. There are two aligned holes punched on opposite ends of the device: a reservoir for feeding solution into the device and a tubing connection port for washing the device with a pump. (FIG. ID) To determine whether single beads can be detected and counted in our bonded device, we loaded a cocktail of antibody-coated beads (10%) and spacer beads into our system. We then added a fluorophore-conjugated secondary antibody to stain and image antibody-coated beads. [0055] During imaging, we observed that we can achieve an average of 10% loading of fluorescently labeled beads into each microwell, and could resolve and count the individual beads. (FIG. IE) Therefore, our technology served well for integrating microbeads into a microwell array and showed the feasibility of detecting individual molecules by counting the number of microbeads within a microwell.

[0056] Off-chip TSA Assay Validation

[0057] To detect scarce protein molecules from a single EV, we employed TSA. The TSA assay relies on the interaction between horseradish peroxidase (HRP) and tyramide. When tyramide reacts with HRP in the presence of hydrogen peroxide, phenolic groups in the tyramide become oxidized, producing tyramide radicals that form covalent bonds with aromatic amino acids rich in electrons. These radicals are then dispersed on the site of amplification, providing a localized signal that is contained to individual beads that can be easily recorded. In this case, the signal does not diffuse away, similar to ELISA. By labeling these tyramide radicals with fluorochrome or biotin, previously undetectable molecules can be observed. Thus, TSA serves well in its application for single EV molecular profiling, given their relatively low abundance of proteins. To demonstrate TSA’s capacity for amplifying fluorescent signals, we targeted EGFR protein from both lysed cell and EV samples of a highly enriched EGFR protein-based cell line (A431).

[0058] We applied a TSA workflow for our TSA off-chip validation. Similarly, we implemented the assay on epoxy group coated beads, which facilitate the antibody coating process on-bead. Because we were targeting EGFR protein from A431 cell and EV samples, an anti-EGFR antibody was coated on the epoxy beads. Because EV lysis conditions remain unstandardized, we optimized the EV lysis conditions with Triton X-100 and sonication. We uncovered no significant differences in the percentage of intact EVs with or without sonication and with 1% or 10% Triton X-100 at 30 minutes of incubation. Thus, we chose 1% Triton X-100 with no sonication. (FIG. 7) By incubating the cell and EV lysate with the antibody-coated beads, we were able to capture EGFR protein and amplify its detection with TSA successfully. (FIG. 2 A)

[0059] Comparatively, with no amplification, there were almost no fluorescent signals from EV samples. Without being bound to any particular theory or embodiment, this may be attributable to the EVs’ relatively low protein abundance compared to cells. Based on the fluorescence intensity, the TSA condition for the EV samples was about 100 folds greater than both the unamplified and negative (no detection antibody) control samples. Similar trends were observed for the cell conditions based on the quantified fluorescence intensity. (FIG. 2B) Therefore, the off-chip validation proved the necessary integration of TSA to resolve undetectable EV protein signals.

[0060] Single EV loading and on-chip TSA

[0061] Single EV loading was optimized in our microwell array. In theory, because our device is occupied with 10,000 microwells, a diluted solution with 1000 EVs should be loaded into the device to achieve single entity loading. However, because our device is primarily occupied by dead space (an area without microwells), a series of titration experiments were performed to compensate for the loss of EVs. Because EVs were not be imaged in our microwell array, so cells were used to parametrize loading into our system. A431 cells, stained with Hoechst, were added at different cell numbers (0, 1000, 2500, 5000, 7500, 10,000, and 12,500) into the microwell array. To achieve digital loading (X = 0.1), the Poisson distribution reports that 9.05% of the wells should be occupied by single entities. Based on imaging, we discovered that after 10,000 cells were added to our device, 10.8% of the wells were occupied with single cells. (FIG. 3 A) In this case, to compensate for dead space loss, 10,000 entities were loaded into the device to achieve the Poisson distribution range.

[0062] With the optimized single-cell loading into our microwells, we then confirmed its translation with EVs. A431 EVs were loaded at different numbers (100, 1000, 10,000, 50,000, and 100,000 EVs) into the microwell array. These numbers were calculated based on NTA measurements after EV purification. Because single EVs were not imaged and detected without amplification, we proceeded with our entire workflow to demonstrate its operability.

[0063] A cocktail of epoxy beads (EGFR capture antibody beads and spacer beads) was added to the device, followed by the EVs. After EV loading into the device, lysis, capture, and amplification was performed. After conducting a series of titrations, we were able to verify that the concentration of 10,000 EVs yielded a number of fluorescent wells (-10%) that closely matched the Poisson distribution range, as well as the reported single-cell loading data. (FIG. 3B) For the 100,000 EV-loaded devices, based on the Poisson distribution (X = 1), 40% of the wells are supposed to remain empty, which is demonstrated by this titration. Further analysis of both the cell and EV titrations are expanded in FIG. 8. With our highly sensitive digital ELISA platform, we demonstrate how we can perform singlemolecule digital detection (FIG. 3C), as well as optimize for achieving single EV loading.

[0064] Single EV PD-L1 molecule detection

[0065] After optimizing the TSA on-chip assay with EGFR protein, we chose to apply our technology for the detection of important cancerous biomarkers, like PD-L1. PD- L1 is an immune checkpoint molecule that plays a significant role in immune evasion in cancer through its interaction with PD-1. Numerous studies have demonstrated that the level of PD-L1 positive(+) EVs in circulation significantly correlates with various aspects of tumor behavior, including size, metastatic status, and therapy response. This presents a new opportunity to develop PD-L1(+) EVs as a cancer biomarker with promising prognostic and diagnostic potential. In this case, there is a pressing need to develop a technology that can accurately and comprehensively profile individual EVs and their PD-L1 expression levels with a high degree of sensitivity. Thus, we applied our highly-sensitive device to quantify the abundance and variability of PD-L1 protein loading in single PD-L1(+) EVs derived from a melanoma cell line (624-mel).

[0066] To validate TSA’s sensitivity and specificity, we detected PD-L1 protein from bulk PD-L1 expressing and PD-L1 knockout 624-mel derived EVs. The TSA off-chip workflow was performed similarly to the EGFR off-chip experiment, with the exception of anti-PD-Ll capture and detection antibodies. Based on the data, the PD-L1(+) EV TSA condition generated a fluorescence signal, while the PD-Ll(-) TSA did not. Because of the EVs ’relatively low protein abundance, there was also no detectable signal for the sample without TSA as well. (FIG. 9)

[0067] Through quantification, the fluorescence intensity for the PD-L1(+) condition was over 20-fold greater than the PD-Ll(-) condition. (FIG. 4A) This indicates the specificity of our TSA workflow and its application in amplifying PD-L1 protein detection. We then transitioned to detecting PD-L1 protein from single EVs in our microwell assay. With the optimized EV-loading numbers reported in the previous section (10,000 EVs for = 0.1), we applied the same workflow for PD-L1(+/-) EVs.

[0068] Based on imaging, we confirmed that the PD-Ll(-) EV loading into the microwells did not generate a signal, while the PD-L1(+) device did. Through quantification, we confirmed that ~8% of our PD-L1(+) microwell array had a signal, which is close to our previously reported single EV loading data. (FIG. 4B) The relative differences in positive well signals can be attributed to differences in PD-L1 expression between single EVs.

[0069] To verify the efficacy of our device in detecting single molecules, we conducted a series of off-chip ELISA studies. First, a standard curve was made by serially diluting recombinant human PD-L1 protein at different concentrations and detecting it with ELISA. Subsequently, PD-L1 expression was measured using ELISA in PD-L1(+/-) EVs at two different loading concentrations (10 and 20 pg/mL). PD-L1 concentration was calculated using the absorbance (A450) and the equation of the standard curve. Between the two EV concentrations, 10 pg/mL of PD-L1(+) EV was found to have 3.65 ng/mL of PD-L1 protein, which was within the range of the standard curve. (FIG. 5 A)

[0070] With this PD-L1 concentration and known EV concentration (particles/mL) measured by NT A, the estimated number of PD-L1 molecules per PD-L1(+) EV was calculated to be between 4-6.5 molecules. (FIG. 5B) This bulk EV measurement is limited to only providing an average number of PD-L1 molecules expressed per EV. On the contrary, our double digital assay can profile individual EVs with their PD-L1 expression. As expected, there was a high variability of PD-L1 protein loading of 1 to 9 molecules per EV. (FIG. 5C) On average, we detected 2.67 molecules/EV in PD-L1(+) EV, comparable to the bulk EV measurement data. (FIG. 5D) Because our device’s average fluorescent bead count is within the range of the ELISA, one can assume that our device is successfully detecting single molecules coming from single EVs.

[0071] To further demonstrate the necessary application of our system for single EV and molecule detection, we performed a series of fluorescent labeling (with anti-PD-Ll targeting antibodies) on bulk PD-L1(+/-) EVs and imaged them with an inverted and superresolution fluorescence microscope. Based on imaging, no differences could be made between the PD-L1(+) and PD-Ll(-) EVs, which indicates that PD-L1 expression is extremely low at the single EV level and our double digital technology is needed to accurately profile key molecules from individual EVs. (FIG. 10)

[0072] To establish the limit of detection (LOD) for our system, we conducted a serial dilution (10⁴, 10³, 10², 10¹, 10°, 10’¹, 10'², 1 O'³) using recombinant human PD-L1 protein and antibody-bound epoxy beads. Subsequently, we deployed the highly sensitive TSA method for signal detection. Our findings revealed a LOD of 0.4 pg/mL, which is >60 times more sensitive than ELISA (LOD = 30 pg/mL). (FIG. 5E) It's noteworthy that our results concur with the existing reports, which describe a LOD ranging from 0.1-1 pg/mL for TSA.³⁸ Altogether, we believe our findings corroborate our device's ability to detect single PD-L1 molecules coming from single EVs, providing a highly sensitive, reliable, and robust platform for accurately profiling significant prognostic and diagnostic cancer markers.

[0073] Conclusion

[0074] The use of EVs as a source of biomarkers for disease diagnosis has generated significant interest. Despite their promise, current diagnostic methods for EVs pose significant challenges. One of the primary issues is the low expression levels of biomarkers carried by EV samples, which make them difficult to detect accurately. Additionally, the physical and biological properties of EVs are complex, further complicating the diagnostic process. As a result, there is a pressing need to develop novel diagnostic techniques that can overcome these challenges and unlock the full potential of EVs as a source of biomarkers. Thus, we have developed a double digital ELISA-based assay that enables the absolute quantification of individual molecules from a single EV. Our approach involves the use of microwells to compartmentalize antibody-coated and uncoated beads evenly, creating a monolayer that is ideal for subsequent digital ELISA. By counting the fluorescent beads within the microwells, our platform enables the precise determination of the expression level of specific biomarkers. With the disclosed technology’s ability to accurately quantify protein abundance and detect biomarkers, the disclosed technology provides a platform useful in the diagnosis and treatment of diseases.

[0075] We employed our integrative microfluidic technology to successfully measure protein abundance coming from single EVs. More specifically, we demonstrated how our technology can measure EGFR molecules coming from single EVs, demonstrating the efficacy of our technique. Not to mention, we demonstrated our ability to achieve the digital loading range for single EVs, which has yet to be demonstrated in microwells. Furthermore, we applied our device to measure the relative abundance of the PD-L1 protein from a melanoma cell line. Based on our research, we have successfully demonstrated the ability of our device to capture and detect individual PD-L1 molecules. Our chip yielded an average of 2.67 fluorescence beads per single EV-positive PD-L1(+) well, while our ELISA data revealed an average of 4-6.5 PD-L1 molecules per individual PD-L1(+) EV. This indicates that our technology is capable of detecting single molecules from single EVs, given that our device is relatively close to the detection threshold. Furthermore, our technology's sensitivity was confirmed by our LOD (0.4 pg/mL), which is 60-fold greater than ELISA. Thus, we have here demonstrated the first-ever double digital detection of single-EV and single-molecule.

[0076] Methods

[0077] Microwell and top-flow chamber fabrication

[0078] The mylar photomasks were designed in AutoCAD and produced through Fineline Imaging. The silicon molds were fabricated at the Singh Center for Nanotechnology at the University of Pennsylvania. The microwell (h = 50 pm) and top flow chamber (h = 100 pm) layers were fabricated with soft lithography using SU-8 3050. The PDMS devices were then bonded together via plasma bonding for 20 s at high power. To increase the hydrophobicity and remove pockets of air bubbles in the microwells, the bonded devices were incubated with Pluronic F-127 (w/v: 0.05%) and degassed for 1 hour at RT, respectively. The devices were then stored in Pluronic F-127 at 4°C until usage.

[0079] Cell culture & EV isolation

[0080] A431 cells were grown in a 150 mm cell culture dish and then expanded to 12 dishes for EV isolation. DMEM (10% FBS, 1% penicillin) was used to culture and passage the cells. After the cells reached confluency, the medium was changed to exosome- depleted DMEM (5% exosome-depleted FBS, 1% penicillin). After 48 hours from media exchange, the collected supernatant was spun at 400g for 5 min and filtered with a 0.22 pm vacuum filter to remove cellular debris. The supernatant was centrifuged twice (Beckman Coulter) at 100,000g for 70 min at 4°C. The EV pellet was resuspended in PBS, aliquoted, and kept at -80°C. PD-L1 (+) and PD-L1 (-) EVs were donated from the Wei Group.

[0081] EV characterization (Qubit, NTA)

[0082] Two different techniques were used to characterize the EVs. Qubit (Thermo Fisher) was used to assess the protein content, and nanoparticle tracking analysis (NTA) was used to determine how many particles were present. Thermo Fisher's protein assay kit was used for Qubit, and measurement was done in accordance with the manufacturer's instructions. The measurement for NTA was carried out at the University of Pennsylvania School of Veterinary Medicine Extracellular Vesicle Core (ZetaView by ParticleMetrix). The analysis employed identical parameters (sensitivity of 75 and shutter of 75).

[0083] PD-L1 antibody [0084] To produce anti-PDLl antibodies (GenScript), synthetic extracellular part (ECD) peptides of PD-L1 were used to immunize mice. Standard ELISA was employed to test different clones of anti-PDLl antibodies for their reactivity to the PD-L1 protein. At least 70 antibody clones were screened, from which pair-matched clones 6G8 and 3F9-Biotin were finally selected for the exosomal PD-L1 ELISA assay. Anti-PDLl antibodies were donated from the Wei Group.

[0085] Antibody biotinylation

[0086] BSA-free antibodies, secondary antibody AF647 (Thermo Fisher; A32787) and cetuximab (anti-EGFR antibody, Selleckchem; A2000) were buffer exchanged to bicarbonate buffer (pH 8.4) using a 40k Zeba column (Thermo Fisher, 87765). The antibody was then incubated for 30 minutes at room temperature (RT) with 20 molar equivalents of biotin-NHS ester (Click Chemistry Tools; B102-1G). Excess biotin-NHS ester linker was then removed using a 40k Zeba column twice. The biotinylated antibodies were then stored at 4°C until usage.

[0087] Bead loading optimization

[0088] Streptavidin-coated magnetic beads (Spherotech; SVM-40-10) were used for bead loading optimization. Before the beads were loaded, they were washed with D.I. water on a PCR magnetic tube rack. To create biotin-NHS-AF647 labeled streptavidin-coated beads, the biotin-NHS-AF647 conjugate was incubated with the streptavidin-coated beads in bicarbonate buffer (pH 8.4) for 30 minutes at RT on a rocker. After incubation, unbound NHS-647F was washed away with PBS. A cocktail of biotin-NHS-AF647 labeled (10%) and spacer (90%) streptavidin-coated beads were then loaded into the device. The beads were allowed to settle for an hour at RT and then centrifuged at 100g for 1 minute to encourage monolayer displacement of the beads. The device was then imaged using an Olympus 1X83 inverted fluorescence microscope.

[0089] Epoxy bead coating

[0090] Epoxy magnetic beads (Spherotech; EM-20-10) were prepared for antibody conjugation. The magnetic beads were washed with D.I. water on a PCR magnetic tube rack. For every 1.6E6 of epoxy beads, 1.25 pg of anti-EGFR and anti-PD-Ll (6G8) antibodies were prepared. The beads and antibody were resuspended into 200 pL of carbonate buffer (pH 9.0), and incubated on a rocker for 20 hours at 37°C. After incubation, the beads were washed and resuspended in PBS. The antibody -bound epoxy beads were then stored at 4°C until usage.

[0091] EGFR and PD-L1 TSA detection off-chip

[0092] For A431 cell lysate, 1.0E6 fresh cells were pelleted and incubated with a IX working concentration of protease inhibitor cocktail (Thermo Fisher; 78430) in 1 mL of RIPA lysis and extraction buffer (Thermo Fisher; 89900) on ice for 15 minutes. After incubation, the solution was centrifuged at 14,000 g for 15 minutes to pellet the cell debris. The supernatant was then transferred and its protein concentration was assessed with Qubit. Protein was then stored at -80°C until usage. For A431 and 624-mel EV lysate, EVs were incubated with 1% Triton X-100 at RT for 30 minutes. Once the lysis was complete, the protein was quantified with Qubit.

[0093] For the TSA protocol, all incubations were performed on a rocker at RT, and all wash steps were performed on a PCR magnetic rack. First, the antibody -bound epoxy beads were incubated with a blocking buffer (2% BSA-PBS) for 30 minutes. After blocking, the beads were washed 3 times with a wash buffer (PBS + 0.1% Tween-20). The beads were then incubated with cell or EV lysate in the blocking buffer for 1 hour. For every 1.6E6 beads (antibody-bound epoxy bead), 250 ng of cell or EV protein was prepared. After protein capture, the beads were washed 3 times with the wash buffer. The biotinylated detection antibody (anti-EGFR or anti-PD-Ll (3F9)) was resuspended in blocking buffer at a concentration of 0.5pg/mL and incubated with the beads for 1 hour. The beads were subsequently washed 3 times with the wash buffer. The beads were then incubated with the streptavidin-HRP (Thermo Fisher; 21130), diluted in 137.5 ng/mL of blocking buffer + 0.1% Tween-20, for 30 minutes. After incubation, the beads were then washed 3 times with the wash buffer. The beads were then incubated with biotin tyramide (Sigma; SML2135) for signal amplification, diluted at a concentration of 0.5pg/mL in 0.1M borate buffer (pH 8.5) + 0.003% H202, for 10 minutes. Once amplification was complete, the beads were washed and then incubated wth streptavidin-647 fluorophore (Biolegend; 405237), diluted in 0.5pg/mL of blocking buffer, for 30 minutes. Finally, the beads were washed and imaged.

[0094] LOD for TSA PD-L 1 detection

[0095] Recombinant human PD-L1/B7-H1 protein (R&D; 156-B7) was used for validating the LOD of TSA. Using anti-PD-Ll (6G8) antibody-bound epoxy beads, a series of different concentrations of recombinant PD-L1/B7-H1 protein was incubated with the beads. The same protocol and reagents as the TSA detection off-chip with PD-L1 (+/-) EVs were applied in this workflow.

[0096] PD-L1 detection with ELISA

[0097] ELISA assay was performed in 96-well plates (Coming® High Bind Microplate, 9018) according to the manufacturer's instructions. Briefly, the plates were coated with 50 pL of capture antibody, the anti-human PDL1 monoclonal antibody, clone 6G8, at a concentration of 5 pg/mL in PBS and incubated overnight at 4°C. After washing the wells 5 times with PBS containing 0.05% Tween-20 (PBST), 200 pL of blocking buffer (1% BSA in PBST) was added to each well, and the plate was incubated for 1 hour at RT.

[0098] Purified 624m el cell line-derived EVs were diluted in PBS to different concentrations, and 100 pL of each dilution was added to the wells in triplicate. The plate was incubated overnight at 4°C with gentle shaking and then washed 5 times with PBST. Biotinylated anti-human PD-L1 monoclonal antibody, clone 3F9, was added at a concentration of 1 pg/mL, and the plate was incubated for 1 hour at RT. After washing the wells 5 times with PBST, streptavidin-HRP (BD Bioscience) was added and incubated for 1 hour at RT.

[0099] Finally, the plate was washed 5 times with PBST and developed with 100 pL of TMB substrate solution. The reaction was stopped with 100 pL of H2SO4, and the absorbance was measured at 450 nm using a microplate reader (BioTek). A standard curve was generated using recombinant human PD-L1 protein (R&D Systems; 156-B7) at concentrations ranging from 0.06 to 4 ng/mL. The PDL1 concentration in the samples was calculated based on the standard curve. The number of PDL1 per EV was calculated based on the molecular weight of PD-L1 ranging from 33-55kD.

[00100] All incubations and washes were performed using an automated plate washer (Fisher Scientific). The data were analyzed using GraphPad Prism software.

[00101] EGFR and PD-L1 TSA detection on-chip

[00102] For each incubation step, the devices are left at RT without agitation or movement. And, for all the wash steps, the devices are washed with 60 pL of wash buffer at a flow rate of 30 pL/hr. For TSA on-chip, a cocktail of antibody-coated (10%) and spacer (90%) epoxy beads were loaded into the microfluidic device. After 1 hour of bead incubation in the device, the device was centrifuged at 100g for 1 minute to create a single monolayer of beads in each well. The device was then loaded with EVs and were allowed to settle for an hour. Subsequently, the device is flowed in with lysis buffer (1% Triton X-100) and then immediately oil (Fluo-oil 7500) to prevent cross-contamination. The device is then incubated with the two-phase solution system for 1.2 hours. After lysis, the detection antibody, followed by the streptavidin-HRP, biotin tyramide, and streptavidin-647 fluorophore were all incubated with the device with the same off-chip protocol and imaged.

[00103] Super-resolution imaging of EVs

[00104] AF488-NHS (Sigma; 41698-1MG-F) was first used to stain all EVs by targeting the surface protein of PD-L1 EVs. EVs (3.4 pg) were mixed with 6 pl of AF488- NHS (1 mM), and the reactions were brought to a final volume of 12 pL with bicarbonate buffer (pH 8.4). After 2 hours of reaction at RT, excess AF488-NHS was removed using a 40k Zeba column twice. Then, stained EVs were introduced to an 8 well-chambered cover glass (Cellvis, CA, USA) and incubated for 30 min to deposit EVs on its surface. The cover glass was then washed with lx PBS and blocked for 30 min by using 1% BSA. Then, 5 pg/ml of primary PD-L1 antibody (3F9-biotin) diluted in 1% BSA was incubated with the sample at RT for 1 hour. After unbound antibodies were washed away, 2 pg/ml of AF647-labeled secondary antibodies (strep-AF647) were incubated with the cover glass for 1 hour at RT. Finally, samples were washed with lx PBS.

[00105] After exchanging IxPBS for imaging buffer (10% glucose, 100 mM Cysteamine, 1% GLOX in PBS)³⁹, dSTORM imaging was acquired on ONI nanoimager (Oxford Nanoimaging, Oxford, UK) equipped with 405 nm, 488 nm, 561 nm, and 640 nm lasers. Two-channel dSTROM data was acquired using the 488 nm and 640 nm laser with a power of 200 mW and 400 mW, respectively, and an exposure time of 20 ms with 1500 frames. Data was drift-corrected and filtered using CODI software (Oxford Nanoimaging) to minimize low-precision and non-specific localization. A diameter between 30 nm and 240 nm and circularity greater than 0.3 were considered EVs. Individual EV clusters and the number of counts were analyzed using CODI software. Note that the number of counts from PD-L1 was taken from dual positive EVs.

[00106] Additional Disclosure

[00107] Extracellular vesicles (EVs) are actively shed by cells existing in healthy and pathological states. They are increasingly recognized as potential circulating biomarkers of disease. To date, areas of precision medicine such as oncology increasingly rely on biopsied tumor tissue. However, sampling beyond the initial biopsy limits tissue immunohistochemistry’s utility in navigating clinical management. Therefore, it is ideal to collect a circulating biomarker (“liquid biopsy”). Notably, EVs exhibit high stability for protecting molecular cargo as well as an abundance of sources. EVs shuttle diverse molecules including proteins and nucleic acids that reflect the composition of their parent cells and may be superior in sensitivity to cell-free DNA (cfDNA) for cancer diagnosis. However, the complex heterogeneity of EVs limits their clinical use as a molecular diagnostic.

[00108] Identifying disease-specific EVs, especially in the earliest stages of cancer development, is notoriously difficult because of the overwhelming background of EVs shed by healthy host cells. Furthermore, the volume of EVs is approximately 10^A6 fold smaller than that of a mammalian cell, limiting the relative abundance of protein molecules that can be analyzed with bulk EV analysis. The stochastic nature of EV biogenesis and the rarity of molecular markers expressed in these nanometer particles renders bulk analysis challenging. Hence, high-throughput single EV or “digital” profiling methods can address these technical gaps to resolve the molecular heterogeneity of EVs to discover meaningful diagnostic markers. A number of single EV techniques have been developed, including fluorescent microscopy of EVs on glass (SEA), small EV flow cytometry, and digital protein platforms. These platforms, however, are limited by their ability to detect or profile the co-expression of rare proteins in a single EV. These limitations include the size detection limit of flow cytometers, the need for large EV samples for single EV fluorescence imaging or flow cytometry, and the multiplexing limit of digital ELISA.

[00109] To address the limitations of current techniques, hydrogel microparticles can serve as a platform for single EV analysis. Hydrogel microparticles have applicability to bioassays, where their porous mesh-like structure allows for biological interactions in three- dimensional space, exhibiting solution-resemble hybridization kinetics. Further, the nonfouling nature of hydrogel materials inhibits nonspecific molecules binding in complex matrices. Additionally, one can load and retain a range of materials - from nanoparticles to mitochondrial DNA to cells - within hydrogel microparticles. Furthermore, various amplification strategies such as polymerase chain reaction (PCR), hybridization chain reaction (HCR), and rolling circle amplification (RCA), can be adapted to hydrogel microparticles.

[00110] Herein, we provide the first use of squeezable hydrogel microparticles as a scaffold to immobilize bulk EVs and conduct an integrated RCA assay for a high-throughput and multiplexed analysis of single EV proteins. This involves the high throughput generation of uniform droplets consisting of methacrylated hyaluronic acid (MeHA) polymer precursor and bulk EVs with a flow-focusing microfluidic droplet generator. It should be understood that the MeHA hydrogel system described herein is illustrative only and does not limit the scope of the present disclosure or the appended claims. EVs are physically arrested within hydrogel microparticles upon UV-based polymerization. Consequently, EVs can be labeled with DNA barcoded antibodies (Ab-DNA) and multiplex protein signals can be amplified by incorporating RCA into the hydrogel microparticle assay for analysis. We validated that MeHA hydrogel microparticles exhibit high compressibility upon physical squeezing, which can align RCA products in a single plane and enable imaging without the need for high- resolution z-stack imaging. Following initial validation and optimization with A431 cell line- derived EGFR lysate and EVs, we performed a multiplexed single EV protein analysis of four different cancer cell lines and profiled the presence/absence and co-expression of cancer markers among single EVs. As a result, we identified that pan-cancer markers are variably expressed among different cancer line-derived EVs. The described method addresses previous shortcomings of other single EV methods that have, for example, relied on single EV isolation in droplets or high resolution single particle imaging on planar substrates. Altogether, we present a high-throughput and multiplexed analysis of single EV proteins for analyzing the molecular heterogeneity of single EVs and defining EV subpopulations that can meet the demands of future clinical biomarker validation studies.

[00111] Results and discussion

[00112] Synthesis and characterization of MHPs

[00113] We first fabricated MHPs by using serial microfluidic techniques including a flow-focusing microfluidic droplet generator for droplet synthesis and a polymerization chamber for UV exposure to induce the UV crosslinking of droplets. MeHA was synthesized by methacryl ati on of hyaluronic acid by esterification with methacrylic anhydride and the degree of substitution was quantified by 'H NMR (Figure 16). A prepolymer solution containing MeHA, EV and a photoinitiator was injected as a dispersed phase into a droplet generator which includes curved structure after pinch-off region for enhanced mixing efficiency (Figure 17). Subsequently, droplets were reinjected into the polymerization chamber to be crosslinked into the hydrogels by photocuring. The height of the polymerization chamber (50 pm) was set to be smaller than the diameter of the droplets to make spheroid-like structures with flat tops and bottoms for accurate measurement of modulus and uniform compression. The optimal concentration of MeHA was found to be 2.5% (w/v), which can reduce the Young’s modulus for squeezing while maintaining the stiffness of the structure.

[00114] To assess the capture efficacy of EVs within the MHPs, 100 nm fluorescent beads were utilized as a model for EVs and tested under various crosslinking conditions, with varying UV exposure times (Figure 2B). The capture efficiency was observed to be higher when the MHPs were synthesized under prolonged UV exposure times, due to the formation of denser networks within the MHPs. Also, the remaining beads within the MHPs were validated after the completion of the assay procedures, including shaking and incubation at elevated temperatures (Figure 2C). A comparison of the results obtained from the initial (To) and final (Ti) assay procedures revealed that MHPs synthesized under low UV exposure times exhibited a significantly higher degree of leakage (60.1% for 0.3 s) than those synthesized under high UV exposure times (3.6% for 2 s). After the whole process, 5.7%, 13.8%, 26.4%, and 34.7% of beads remained for MHPs synthesized at 0.3, 0.6, 1, 2 seconds of UV exposure respectively, and we chose 1 second as an optimal UV exposure time to enhance the permeability of MHPs (data not shown).

[00115] We further characterized mechanical properties of MHPs using atomic force microscopy (AFM). In addition to MHPs, PEG-based hydrogel microparticles (PHPs) were fabricated for the comparison with the MHPs. Nanoindentation of MHPs and PHPs revealed significant difference in the modulus of the two groups and decrease of MeHA concentration leads to the decrease of modulus (Figure 2D, E). Subsequently, the compressibility of MHPs and PHPs was evaluated by applying a force (-960 pN) to the particles. As expected, MHPs demonstrated high compressibility (Figure 2E, F), whereas PHPs were unable to be compressed (Figure 18). The diameter of MHPs increased by a factor of two, resulting in a reduction in height from 50 pm to 12.5 pm. The reduced height is comparable to the z-axis resolution of the 20x objective lens (-7.4 pm), calculated by Abbe’s diffraction formula when assuming a wavelength of 569 nm. This indicates that MHPs can be imaged in a single plane without the need for z-stack imaging.

[00116] Optimization of RCA

[00117] RCA was been integrated into the hydrogel microparticles due to its unique features, which can be performed on a solid support or inside the complex structures such as hydrogels. Further, the amplicons generated by RCA are anchored on a structure with a size of hundreds of nanometers to a few micrometers, which enables retention of amplicons within the hydrogel networks and allows for digital counting of dot numbers.

[00118] The workflow to perform the RCA-based assay is represented as follows (Figure 3 A). The initial step involves the capture of target proteins by DNA-labeled antibodies. This is followed by the introduction of a padlock probe, which is designed to hybridize with the DNA. Then, the padlock probe is circularized by DNA ligase, followed by the introduction of Phi29 polymerase, which synthesizes long strands of DNA based on the sequence of the padlock probe. This long single-stranded concatemer of DNA is tagged by multiple numbers of fluorophore-labeled DNA, resulting in an amplification of the signal.

[00119] In order to optimize RCA using MHPs, we targeted epidermal growth factor receptor (EGFR) from lysates of A431 cells, which is a highly enriched protein-based cell line. Protein molecules from cell lysates were loaded into the droplets during the fabrication process and conjugated to the remaining unreacted methacrylate groups via azaMichael addition reaction (Figure 19). We first varied the antibody -DNA (Ab-DNA) concentration prepared by TCO/Tz chemistry (see Methods section for details) to minimize the background dots and selected the Ab-DNA concentration of 62.5 ng/ml, which maximized the signal-to-noise ratio (Figure 3B). In addition, the optimal reaction time for RCA was determined (Figure 3C). To verify the consistency of the reaction without the formation of background RCA dots throughout the RCA process, the number of RCA dots was measured at each time point. It was confirmed that the number of RCA dots remained consistent, while the fluorescent intensity from MHPs increased according to the reaction time. The optimal reaction time was set to be 3 hours, at which point the fluorescence intensity of the RCA dots was sufficiently strong to be analyzed in the imaging conditions (800 ms of exposure time). We further validated that the RCA dots were generated only when all assay components were present (Figure 3D, E)

[00120] EV analysis by MHPs

[00121] After optimizing RCA using MHPs, our technology was applied to the analysis of EVs. EVs derived from A431 cells were selected as the model EV, with EGFR markers serving as the target. A series of titration studies demonstrated that the LoD was 28 EVs/hydrogel (Figure 4A), which is in accordance with the predictions made on the basis of the capture efficiency of MHPs and the prevalence of EGFR in A431 EVs. Subsequently, four distinct probe sets (barcode/PLP/fluorescently labeled DNA) were designed to expand the multiplexing capability of our technology. The specificity of the probe set was assessed by modifying the anti-EGFR antibody with four distinct barcodes and reacting with each PLP/fluorescently labeled DNA set (Figure 4B, C). It should be noted that the cell lysate was used for the probe set specificity test to avoid the heterogeneity of signals associated with EVs. The results confirmed two important factors for multiplexing; 1) high specificity with negligible nonspecific dots; and 2) similar reactivity between probe sets. To evaluate the specificity of our technology, a control experiment was performed to compare the number of dots obtained from no EV control and isotype control (anti-IgG isotype antibody) to those obtained with target-specific antibodies (anti-EGFR antibody) (Figure 4D). The number of RCA dots from both the no EV control and the isotype control was considerably lower than that of the positive control, indicating that the RCA dots are generated in a highly specific manner.

[00122] We then demonstrated multiplex detection of A431 EVs by targeting three surface markers of EVs; EGFR, EpCAM and CD9. EGFR and EpCAM was chosen as pancancer marker and CD9 was chosen as an EV marker which is known as the most prevalent marker among the tetraspanins. To analyze the single-positive, double-positive and triplepositive EVs, we developed a customized analysis software based on the CellProfiler. The customized pipeline enables segmentation and counting of RCA dots from different fluorescent channels, which leads to the automation of the analysis process (Figure 20). Note that the single RCA dot was considered as a single EV, regardless of the size of the RCA dot. To account for background RCA dots, the same number of EVs from fetal bovine serum were used as a control. As the background dots of fetal bovine serum from each marker are not overlapped, the number of background dots was subtracted solely from the single-positive dots. Applying this strategy, we analyzed distribution of three markers in A431 EVs, and revealed that 73.1%, 25%, and 1.9% of EVs were single-positive, double-positive and triplepositive, respectively (Figure 4E, F). Most of the double-positive EVs originated from CD9 and EGFR colocalization, which is the combination of prevalent pan-cancer and EV markers. Additionally, a comprehensive single EV map was constructed across three markers to achieve single EV resolution (Figure 4G).

[00123] To date, glass or gold surfaces have been frequently used as a substrate for single EV analysis methods. Although these methods demonstrated their utility in single EV detection, we speculated that the numbers from the surface assay might underestimate the populations of single EVs due to the steric restrictions of the two-dimensional surface. Therefore, we tested the co-expression ratio of CD9 and EGFR from A431 EVs using surface assay and EV flow cytometry (Figure 4H). EV flow cytometry was employed to explore an ideal solution assay in which no restrictions were applied during the assay. Interestingly, there was a significant difference in co-expression ratio between the surface assay and solution assay with the results of the hydrogel assay falling between these ratios. This is because the hydrogel provides biological interactions in three-dimensional space, thus minimizing steric hindrance during the assay. In this study on co-expression ratios, we demonstrated that hydrogel-based single EV detection is more accurate than surface-based assays in terms of identifying marker expression. The use of a hydrogel microparticle assay thus not only addresses sensitivity by enabling larger surface area of interactions with epitopes, but also provides a scaffolding for enzymatic amplification via RCA of protein signals.

[00124] Profiling cancer cell EV protein marker heterogeneity by MHPs

[00125] A number of pan-cancer markers have been identified for different cancers, but the clinical utility of a single marker is limited due to insufficient sensitivity or specificity. Following preliminary validation of the specificity and multiplexing, we proceeded to perform a 4-plex analysis of single EVs among other cancer cell lines: A431 (epidermal), A549 (lung), MCF7 (breast) and PANCI (pancreatic). To first calibrate and validate the MHP assay, we investigated the expression of pan-cancer markers EGFR, EpCAM, and MUC1 as well as exosomal marker CD9 between parent cells and EVs by fluorescent intensity measurements from cell immunostaining and number of RCA dots per gel from the MHP assay, respectively. As expected, it was found that distinct cell lines expressed an abundance of each cancer marker as reported in literature: A431 cells express EGFR; PANCI cells express MUC1; and MCF7 cells express EpCAM (Figure 5 A). Interestingly, it was observed that EV and whole cell CD9 measurements reflected an inverse relationship in comparison to pan cancer marker expression between whole cells and shed EVs. Based on these observations, CD9, a non-cancer marker, was excluded from the correlation analysis of pan-cancer markers between parent cells and EVs. Furthermore, each cell line’s single EV counts reflect similar cancer protein expression profiles from whole cells with our data showing good concordance (Spearman correlation coefficient r = 0.84) (Figure 5A, B). In addition to constructing comprehensive single EV maps that illustrate the heterogeneity of protein marker presence/absence for each cancer line, we quantified the distribution of protein marker co-expression (Figure 5C) for different marker combinations. It was determined that less than 1% of EVs from different cancer lines express all markers. As expected, EV CD9 counts were relatively high in abundance (35% or higher) since tetraspanin markers such as CD9 should be constitutively expressed among EVs.

[00126] To demonstrate that the degree of multiplexing is not restricted by spectral limitations, we validated that the MHP assay is compatible with cyclic imaging. This was demonstrated with CD9 and EGFR signals, such that fluorophore DNA following the first round of hybridization could be stripped from RCA products by formamide-based denaturation within 10 min. As a result, future studies can incorporate iterative rounds of fluorophore DNA hybridization/stripping, enabling higher order multiplexing for single EV analysis. Altogether, these findings demonstrate our MHP assay’s performance to sensitively detect and profile the co-expression of protein molecules in a high throughput manner from single EVs, providing a highly sensitive, reliable, and robust platform for accurately profiling significant EV subpopulations and prognostic cancer markers.

[00127] Discussion

[00128] The promise of analyzing EVs for molecular disease diagnostics has produced significant interest. However, current diagnostic tools for EVs are significantly limited by the rarity and compositional heterogeneity of protein biomarkers expressed on EVs. There is an unmet need to develop accurate and sensitive techniques that address these challenges to accurately validate the clinical utility of EV biomarkers. The work presented here, a MHP assay integrated with RCA, enables a high-throughput, multiplexed, and ultrasensitive analysis of single EV proteins. Our approach addresses previous shortcomings of other single EV methods such as single particle isolation in droplets, multiplexing and steric hindrance limitations of two-dimensional single EV imaging, as well as the cost of small particle flow cytometry tools with expensive optics. As a result, image acquisition is cheap, simple, and fast; and EV analysis was made high throughout as a single image contains > 10^A3 EV and an automated pipeline to describe the presence/absence and coexpression of protein markers for single EVs for different cancer cell lines. This sensitive platform represents a promising path for defining the earliest diagnosis of disease development. [00129] In addition to the described system, one can also customize the backbone of MHPs to be decorated with capture antibodies conjugated to chemical linkers (e.g., streptavidin-biotin) to isolate cell-type specific EVs without direct mixing in microfluidics and analyze for validating unique protein expression patterns for disease diagnosis. Further, although a 4-plex analysis was performed, we have demonstrated that the disclosed technology is not constrained by spectral limitations. We validated that cyclic imaging can be integrated with RCA products to exponentially increase the degree of multiplexing that can be achieved. As a result, a detailed single EV analysis allows for the classification of rare EV subpopulations and enable organ-of-origin identification of cancer development from obtainable clinical biofluids such as plasma. Such methods that enable a multiplexed profiling of single EVs provides insight to protein patterns for developing better strategies for studying EVs and resolving the clinical limitations of early cancer detection.

[00130] Methods

[00131] Fabrication of microfluidic device

[00132] The device was first designed using a design software (AutoCAD, USA) and 50 pm height of SU-8 master mold was fabricated by a conventional photolithography process. To make the droplet generator, a Poly dimethyl siloxane (PDMS; Corning, USA) mixture (10:1 base to curing agent ratio) was prepared, poured over the master mold, and cured the polymer (70 °C, overnight). The cured PDMS slab was then peeled off, cut into the desired size, and punched holes (0.8 mm biopsy punch; Miltex, USA) to define fluidic ports. As a fluidic bottom substrate, a PDMS mixture was coated onto the glass slides and partially cured (60 °C, 2 h). The fluidic device was assembled by attaching the PDMS slab to the partially cured surface of the glass slides. The final device was cured (60 °C, overnight) to ensure leak-tight bonding. To make the polymerization chamber, 1.6% (w/w) of the iron powder (Thermo Fisher Scientific, USA) was mixed with a PDMS mixture to prevent scattering of UV inside the channel.

[00133] MeHA synthesis

[00134] Hyaluronic acid (HA, 90kDa; Lifecore Biomedical) was methacrylated by esterification with methacrylic anhydride (MA). Briefly, HA (5.0 g) was dissolved in deionized (DI) water (500 mL) in a three-neck round bottom flask. The reaction was cooled on ice and the pH adjusted to 8.5 by the addition of IN NaOH prior to the addition of MA (11.14 mL). The reaction was vigorously stirred, maintaining pH 7.5-8.5 by the addition of 1N NaOH for 3 h before allowing the solution to stir at room temperature overnight. Purification was performed by dialysis against DI water for 10 days (6-8kDa MWCO; Spectra/Por). The final product was frozen at -80°C, lyophilized, and characterized by 1H NMR (500 MHz; Varian Unity Inova). The degree of substitution was quantified by determined to be 30%.

[00135] Hydrogel microparticle fabrication and squeezing

[00136] A precursor solution was prepared by mixing 2.5% (w/v) of MeHA dissolved in triethanoamine buffer (0.2 M, pH 8; Sigma Aldrich, USA) with 0.2% (w/v) Lithium Phenyl-2,4,6-trimethylbenzoylphosphinate (LAP; Takara, Japan) for MHP and 20% (v/v) of PEG700DA (Sigma Aldrich), 40% (v/v) of PEG600 (Sigma Aldrich) and 0.2% (w/v) LAP for PHP. Precursor was injected into the microfluidic device as a dispersed phase using a syringe pump (New Era Pump Systems Inc., USA) while HFE-7500 with 2% (w/w) fluorosurfactant (RAN biotechnologies, USA) was injected as a continuous phase. The generated droplets were reinjected to the polymerization chamber for the gelation through UV exposure (10 mW/cm2). After the gelation, lH,lH,2H,2H-perfluoro-l -octanol (Sigma Aldrich) was introduced to remove the surfactant and IxPBST (IxPBS containing 0.05% (v/v) tween 20) was added for the phase transition from oil to aqueous phase. Finally, MHPs were washed three times in IxPBST. To squeeze MHPs, a cover slip (22x 22mm; Thermo Fisher Scientific) was put on top of MHPs which were mounted on a glass slide.

[00137] MHP capture efficiency measurements

[00138] 2% (v/v) of Cy5-labeled 100 nm nanoparticles (Nanocs, USA) were encapsulated in MHPs with varying UV exposure times. Fluorescent images of MHPs were taken prior to transfer to the aqueous phase. Following transfer to the aqueous phase, MHPs were washed three times in IxPBST and fluorescent images were taken to measure the remaining nanoparticles. The capture efficiency was calculated by dividing the fluorescent intensity of the MHPs in the aqueous phase by the fluorescent intensity of the MHPs in the oil phase. To assess the leakage of nanoparticles during the assay procedure, MHPs were incubated at 25 °C for 1 hr and then incubated at 37 °C for 3 hr in a thermomixer (1500 rpm). Subsequently, fluorescent images of the MHPs were taken and the fluorescent intensity of the initial and final assay procedures was compared.

[00139] Atomic Force Microscopy (AFM) measurements [00140] Young’s modulus of MHPs and PHPs were measured using the Asylum AFM (Oxford Instruments, USA). MHPs mounted on a glass slide was fixed by Cell-Tak (Corning) and submerged in PBS for measurements. A spherical tip indenter, having a tip radius of 5 pm, was calibrated (spring constant 0.047 N/m) to probe the MHPs and PHPs to an indentation depth of 0.1 -0.6 pm. Young’s modulus of the sample was obtained by fitting the force-indentation depth data to a Hertzian contact model.

[00141] where F is the force, E is the Young’s modulus, v is the Poisson’s ratio, R is the radius of the indenter tip and 5 is indentation depth. Poisson ratio was set to 0.5.

[00142] EV isolation from cell culture

[00143] A431, A549, MCF-7, and PANC-1 cell lines were purchased from the American Type Culture Collection. All cell lines were maintained in Dulbecco’s modified Eagle’s medium, supplemented with 10% fetal bovine serum, 100 IU of penicillin, and streptomycin (100 pg/ml). Cells were grown in a 150 mm cell culture dish and subsequently expanded to 12 dishes for EV isolation. DMEM (10% FBS, 1% penicillin) was used to culture and passage the cells. After cells reached confhiency, media was changed to exosome-depleted DMEM (5% exosome-depleted FBS, 1% penicillin). After 48 h from media exchange, the collected supernatant was spun at 400 g for 5 min and filtered with a 0.22 pm vacuum filter to remove cellular debris. The supernatant was centrifuged twice (Beckman Coulter, NC9146666) at 100,000 g for 70 min at 4 °C. The EV pellet was resuspended in PBS and further purified using size exclusion chromatography (70 nm qEV column, Izon science). EVs were kept as aliquots at -80 °C before use.

[00144] EV characterization

[00145] Following EV isolation, samples were characterized in two different ways. The protein concentration was measured using Qubit (Thermo Fisher) and the number of particles was calculated using nanoparticle tracking analysis (NTA). For Qubit, the protein assay kit (Thermo Fisher) was used and the company protocol was followed for measurement. For NTA, the measurement was carried out using the ZetaView PMX220 Twin instrument (Particle Metrix) at the University of Pennsylvania School of Veterinary Medicine Extracellular Vesicle Core. The analysis employed the following parameters: sensitivity of 65 and shutter of 100.

[00146] EGFR detection from A431 cell lysate [00147] 5.0E6 A431 cells were pelleted and treated with a IX working concentration of protease inhibitor cocktail (Thermo Fisher, 78430) in 0.5 mL of RIP A lysis and extraction buffer (Thermo Fisher, 89900) on ice for 15 min. After incubation, the solution was centrifuged at 14,000 g for 15 min to pellet the cell debris. The supernatant was then transferred, and its protein concentration was assessed with Qubit. Protein was then stored at -80 °C until use.

[00148] Antibodies

[00149] Cetuximab (anti-EGFR antibody, Selleckchem, A2000), anti-EpCAM antibody (Bioxcell, BE0386), anti-MUCl antibody (Biolegend, 355602), and anti-CD9 antibody (Biolegend, 312102). All antibodies were tested on positive cell lines and validated before use.

[00150] Cell immunostaining

[00151] All cell lines were individually passaged into an 8-chambered coverglass system (Cellvis, C8-1.5P) and grown to confluence prior to fixation with 4% PFA-PBS). Cells were blocked with 2% BSA-PBS to reduce background signals. Then, cells were successively treated with primary antibodies against EGFR, EpCAM, MUC1, and CD9 (5 pg/mL prepared in BSA-PBS) and Alexa Fluor 647 secondary antibodies (2 pg/mL prepared in BSA-PBS), Goat anti -Human IgG and Donkey anti -Mouse IgG, with 4 wash steps with PBS between and after antibody labeling steps to remove excess antibody solution.

[00152] DNA barcodes and probes

DNA barcode and probe sequences used in this work are included in Table 1 below:

[00153] DNA barcodes were designed to have a unique sequence for padlock probe hybridization. Padlock probes were encoded with sequences that hybridize to fluorophore dye-linked probes that serve as a barcode for each marker of interest. For multiplexing, FAM, Cy3, and Cy5 fluorophores were attached to the probes. [00154] Antibody-DNA conjugation

[00155] BSA free antibodies were buffer exchanged to PBS-bicarbonate buffer (lOOmM sodium bicarbonate in PBS, pH8.4) using a 40k Zeba column (LIFE Technologies, A57758). The antibody was incubated with 15 molar equivalents of TCO-PEG4-NHS Ester (Vector Labs, CCT-A137) for 25 mins at RT after which unreacted TCO-PEG4-NHS Ester was removed using two successive 40k Zeba column buffer exchanges. Degree of labeling (DOL) was checked by incubating antibodies with 10 molar equivalents of Cy3 Methyl Tetrazine (Vector Labs, CCT-1018) for 25 mins at RT (10% DMSO) before any remaining Cy3 Methyl Tetrazine was removed using two successive 40k Zeba column buffer exchanges. Cy3 Antibody ratio was measured using the Nanodrop UV/Vis mode (Thermo Scientific) at A550/A280 and calculated from the known extinction coefficients of the dye (150,000 M-ls-1, CF280 0.05) and protein (215,000 M-ls-1).

[00156] ImM of amine-modified DNA oligo (Integrated DNA Technologies) was buffer exchanged to PBS-bicarbonate buffer (lOOmM sodium bicarbonate in PBS, pH8.4) using a 7k Zeba column (Thermo Fisher, 89878). The DNA oligo was incubated with 10 molar equivalents of Methyl Tetrazine-PEG4-NHS Ester (Vector Labs, CCT-1069) for 25 mins at RT (10% DMSO), after which excess Methyl Tetrazine-PEG4-NHS ester was cleared using three successive 7k Zeba column buffer exchanges. Tz:DNA ratio was calculated from Nanodrop UV/Vis measurements at A520/A260 and the known extinction coefficients of the tetrazine (438 M-lcm-1) and DNA (as supplied by the manufacturer). Measurement at two different dilutions was required given the much stronger molar absorbance of the DNA. TCO-labeled antibody and Tz-labeled DNA were mixed with appropriate DNA stoichiometry (Cy3 Antibody ratio minus 0.5, such that the TCO-antibody sites are in slight excess) and incubated for 45 mins at RT.

[00157] Small EV Flow Cytometry

[00158] EVs were labeled with 10 pg/mL of primary antibody cocktail of anti- EGFR and anti-CD9 in 2% BSA-PBS for 1 hr and purified by size exclusion chromatography (70 nm qEV single column, Izon science) to remove unlabeled antibodies. Single use qEV columns were used to remove excess primary antibodies and 70 nm columns were used to include the 70-1000 nm EV range. After the EV solution was loaded, PBS was used to collect 700 pL of dead volume. The dead volume tube was discarded and 510 pL of eluate was collected in PBS to achieve a pure EV population. EVs were then concentrated using a lOOkDa Amicon concentrator device (Millipore Sigma, UFC510096). Primary Ab-labeled EVs were treated with 2 pg/mL of secondary antibody cocktail (Goat anti-Human IgG, Alexa Fluor 647 for anti-EGFR and Donkey anti-Mouse IgG, Alexa Fluor 488 for anti-CD9) in 2% BSA-PBS for 30 mins and subsequently purified with a 70 nm qEV single column. The analysis of EVs was carried out using the BD FACSymphony Al Cell Analyzer from the Penn Cytomics & Cell Sorting Shared Resource Laboratory.

[00159] Unreacted methacrylate groups characterization

[00160] To identify the presence of remaining methacrylate groups in MHPs, 50 pL of FITC-PEG-SH (10 mg/ml) was mixed with 50 pL of MHPs in IxPBST and incubated at 37 °C for 4 hr in a thermomixer (1500 rpm). After incubation, MHPs were washed 3 times in wash buffer (IxPBST). For blocking study, 50 pL of PEG-SH (10 mg/ml) was used and reacted under the same conditions as those employed for FITC-PEG-SH. For the conjugation of protein molecules, 10 pL of cell lysates extracted from A431 cells were mixed with 90 pL of MHPs in TEOA buffer and incubated at 25 °C for 12 hr in a thermomixer (1500 rpm). After incubation, MHPs were washed 3 times in wash buffer (IxPBST).

[00161] Rolling Circle Amplification assay in MHPs

[00162] The assay composed of four different steps; immunolabeling, ligation, amplification and fluorescent labeling. For immunolabeling, 50 pL of Ab-DNA (0.125 pg/mL) was prepared in IxPBS with 5% BSA and mixed with 50 pL of MHPs in IxPBST. Then, incubation was conducted at 25 °C for 30 min in a thermomixer (1500 rpm). After incubation, MHPs were washed 3 times in wash buffer (IxPBST). For ligation step, ligation mixture was prepared by combining 20 nM PLP, 400 mU/pL T4 DNA ligase (New England Biolabs, USA) in lx T4 DNA ligase buffer (New England Biolabs). The 50 pL of mixture was then combined with 50 pL of MHPs and incubated at 37 °C for 30 min. After three washes, 800 mU/pL of phi29 polymerase (Bioresearch Technologies, England), 250 pM dNTPs (New England Biolabs) in lx phi29 DNA polymerase reaction buffer was prepared for amplification step. The 50 pL of mixture was combined with 50 pL of MHPs and incubated at 37 °C for 3 hr. Finally, after three washes, 50 pL of mixture containing 250 nM fluorescent probe in probe hybridization buffer (Molecular Instruments, USA) was combined with 50 pL of MHPs and incubated at 37 °C for 20 min. All incubation steps were conducted with shaking at 1500 rpm.

[00163] Assay in the microwell [00164] Clear scratch and UV-resistant acrylic sheets (12" x 12" x 1/8”) were used for microwell fabrication. Before laser cutting, double-sided sticky tape was placed onto one side of the acrylic for glass bondage. A PLS4.75 laser system was then used to cut a 6 mm diameter hole into the acrylic sheet. After cutting, the acrylic was bonded to a cover glass slide. EVs were incubated in a microwell for 30 min at RT to settle down on the surface of the glass. After incubation, microwells were washed 3 times in wash buffer. Rest of the step follows the same protocol of assay in MHPs except for the incubation at 37 °C using an incubator.

[00165] Image acquisition and analysis

[00166] Both MHPs and microwells were imaged using an 1X83 inverted fluorescence microscope (Olympus, Japan). For imaging MHPs and immunostained cells, 20x objective was used and for microwells, 40x objective was used. The fluorescence intensity of MHPs and cells were analyzed with the aid of FIJI software (National Institute of Health, USA). The total number of dots and colocalization between dots from a different fluorescent channel was analyzed by Cell Profiler with customized pipelines specific for recognizing and counting dots.

[00167] Aspects

[00168] The following Aspects are illustrative only and do not limit the scope of the present disclosure or the appended claims. Any part or parts of any one or more Aspects can be combined with any part or parts of any one or more other Aspects.

[00169] Aspect 1. A method, comprising: releasing at least one protein carried by an extracellular vesicle (EV) while the EV is in a container free of other EVs; and collecting a signal associated with the at least one protein.

[00170] Aspect 2. The method of Aspect 1, further comprising preventing intermixing between the protein released from the EV and protein released from other EVs. This can be accomplished by, for example, placing oil on top of the proteins released from a particular EV to prevent those proteins from mixing with proteins released from a different EV.

[00171] Aspect 3. The method of any one of Aspects 1-2, further comprising immobilizing the at least one protein. [00172] Aspect 4. The method of Aspect 3, wherein the at least one protein is immobilized to a first antibody that binds to the at least one protein. A protein can also be immobilized to, for example, an aptamer.

[00173] Aspect 5. The method of Aspect 4, wherein the first antibody is associated with a capture microbead.

[00174] Aspect 6. The method of Aspect 5, wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead that is free of the first protein or essentially free of the first protein. A spacer microbead can be, for example, a microbead that is not coated with antibodies; such a microbead does not capture proteins and can then serve as a spacer only.

[00175] Aspect 7. The method of any one of Aspects 1-6, further comprising generating the signal.

[00176] Aspect 8. The method of Aspect 7, wherein the signal is generated by tyramide signal amplification. Other methods for generating the signal include, for example, rolling circle amplification (RCA) and other enzymatic methods used in ELISA approaches.

[00177] Aspect 9. The method of any one of Aspects 1-8, further comprising classifying the EV based on the signal.

[00178] Aspect 10. The method of any one of Aspects 1-9, further comprising classifying, based on the signal, a source of the EV.

[00179] Aspect 11. The method of any one of Aspects 1-10, further comprising selecting a therapeutic treatment based on the signal.

[00180] Aspect 12. A method, comprising: partitioning a plurality of EVs among a plurality of containers such that a given partitioned EV is in a container free of other EVs; releasing proteins from at least some of the partitioned EVs; collecting signals from the released proteins.

[00181] Aspect 13. The method of Aspect 12, wherein the partitioning comprises contacting the plurality of EVs to a plurality of containers under conditions that give rise to partitioned EVs.

[00182] Aspect 14. The method of Aspect 13, wherein the plurality of EVs is distributed among the plurality of containers according to a Poisson distribution with = 1. [00183] Aspect 15. The method of any one of Aspects 12-14, wherein a container has disposed therein (i) capture microbeads configured to bind to released proteins and (ii) spacer microbeads configured to be free of binding to released proteins.

[00184] Aspect 16. The method of Aspect 15, wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

[00185] Aspect 17. The method of any one of Aspects 12-15, further comprising generating the signal.

[00186] Aspect 18. The method of Aspect 17, wherein the signal is generated by tyramide signal amplification.

[00187] Aspect 19. The method of any one of Aspects 12-18, further comprising classifying a partitioned EV based on a collected signal.

[00188] Aspect 20. The method of any one of Aspects 12-19, further comprising classifying a source of the partitioned EV based on a collected signal. Such a source can be, for example, a cell, a patient, and other sources of EVs. In this way, patients who exhibit EVs with one set of characteristics can be classified as being part of a first population, and patients who exhibit EVs with a different set of characteristics can be classified as being part of a second population.

[00189] Aspect 21. The method of any one of Aspects 12-20, further comprising selecting a therapeutic treatment based on a collected signal.

[00190] Aspect 22. A system, comprising: a plurality of containers, the containers optionally comprised in a substrate; and a container having disposed therein (i) capture microbeads configured to bind to at least one protein released by an EV and (ii) spacer microbeads configured to be free of binding to the at least one protein, and wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

[00191] Aspect 23. The system of Aspect 22, wherein a container is a microwell, the microwell optionally being hexagonal in profile. A container can have a cross-sectional dimension (which can be, for example, a height/depth, a side length, or even a width) in the range of a few microns to tens of microns (such as from 1 micron to about 100 microns), in some example embodiments. As but one example, a hexagonal microwell can have a side length of from 20-25 pm, a width of 40-50 pm, a depth of 45-55 pm. The foregoing ranges are exemplary only. [00192] Aspect 24. The system of Aspect 22, wherein a capture microbead comprises an antibody complementary to the at least one protein released by the EV.

[00193] Aspect 25. The system of any one of Aspects 22-24, further comprising an imager configured to collect a signal associated with a protein released by an EV in a container.

[00194] Aspect 26. A method, comprising: with an EV immobilized in a matrix, collecting a signal indicative of a labeled probe molecule being bound to a pendant nucleic acid associated with the EV. As an example, a pendant nucleic acid may be conjugated to an antibody to form a pendant molecule. In another example, a label probe molecule may be bound to a pendant molecule associated with an EV.

[00195] A matrix can be a hydrogel, for example, although other pervious materials besides hydrogels can be used. A matrix can be selected on the basis of being suitable - such as in terms of pore size - to immobilize an EV but also allow diffusion of reagents to and from the immobilized EV. As an example, a hydrogel can be selected to immobilize an EV while also allowing performance of RCA on the immobilized EV.

[00196] Aspect 27. The method of Aspect 26, wherein the matrix comprises a hydrogel. A hydrogel can comprise, for example, MeHA. Non-MeHA hydrogels are also considered suitable.

[00197] Aspect 28. The method of any one of Aspects 26-27, wherein the labeled probe molecule comprises a fluorophore conjugated to a probe nucleic acid. Non-limiting FIG. 13 depicts such a configuration.

[00198] Aspect 29. The method of Aspect 28, wherein the probe nucleic acid is complementary to the pendant nucleic acid. Non-limiting FIG. 13 depicts such a configuration.

[00199] Aspect 30. The method of any one of Aspects 26-29, wherein the pendant nucleic acid is conjugated to an antibody. Non-limiting FIG. 13 depicts such a configuration. An antibody can be selected on the basis of the antibody’s affinity for a protein of interest.

[00200] It should be understood that the disclosed technology contemplates the use of multiple antibody -nucleic acid conjugates. For example, a user can use a first antibody- nucleic acid conjugate that includes an antibody having an affinity for a first protein of interest and also a second antibody-nucleic acid conjugate that includes an antibody having an affinity for a second protein of interest. In this way, a user can screen an EV for multiple proteins of interest, by different conjugates that have affinities for the different proteins of interest.

[00201] Aspect 31. The method of Aspect 30, wherein the antibody binds a protein associated with the EV. Non-limiting FIG. 13 depicts such a configuration.

[00202] Aspect 32. The method of any one of Aspects 26-31, further comprising compressing the matrix. As described herein, the matrix can be compressed while the signal is collected. Without being bound to any particular theory or embodiment, compression converges signal -bearing complexes toward a single z-plane, thereby allowing a user to more effectively visualize the signal-bearing complexes that may be present. This is depicted by FIG. 11C, which shows the convergence of signal -bearing complexes toward a common z- plane.

[00203] Aspect 33. The method of Aspect 32, further compressing the matrix by at least 10% in a direction. For example, the matrix may be compressed by about 50%. For example, a matrix that has an uncompressed thickness of 20 pm can be compressed to a thickness of 18 pm. A matrix can be compressed to a thickness of, for example, from about 160 pm to about 80 pm, from about 5 to about 50 pm, from about 6 to about 18 pm, from about 8 to about 14 pm, or even from about 10 to about 12 pm.

[00204] Aspect 34. The method of any one of Aspects 26-33, further comprising effecting rolling circle amplification (RCA) on the pendant nucleic acid. Other nucleic acid amplifications techniques can also be performed, and it should be understood that RCA is not the exclusive amplification technique that can be performed.

[00205] Aspect 35. A method, comprising: immobilizing an EV within a matrix; contacting the EV with an antibody that binds to a protein associated with the EV, wherein the antibody is conjugated to a pendant nucleic acid; amplifying the pendant nucleic acid; contacting the pendant nucleic acid with a probe comprising a fluorophore conjugated to a probe nucleic acid, the probe nucleic acid being complementary to the pendant nucleic acid; collecting a signal indicative of the probe being bound to the protein.

[00206] Aspect 36. The method of Aspect 35, wherein the matrix comprises a hydrogel.

[00207] Aspect 37. The method of any one of Aspects 35-36, wherein the EV is immobilized through UV crosslinking. An example of such immobilization is described elsewhere herein; as shown, EVs can be disposed in a hydrogel precursor, which precursor is then crosslinked to immobilize the EV.

[00208] Aspect 38. The method of any one of Aspects 35-37, wherein the matrix is compressed, the matrix optionally being compressed to a height of less than about 6 pm, less than about 8 pm, less than about 10 pm, less than about 12 pm, or less than about 14 pm.

[00209] Aspect 39. The method of any one of Aspects 35-38, wherein the matrix comprises at least about 1%, at least about 2%, at least about 2.5%, at least about 3%, at least about 4%, at least about 5%, at least about 6%, at least about 7%, at least about 8%, or at least about 9% methacrylated hyaluronic acid (MeHA).

[00210] Aspect 40. The method of any one of Aspects 35-39, wherein from about 20 EVs to about 36 EVs are immobilized within a matrix. For example, 22 EVs may be immobilized within a matrix. In another example, 28 EVs may be immobilized within a matrix. Such a matrix can be present as a bead or other particle.

[00211] Aspect 41. The method of any one of Aspects 35-40, wherein the signal is collected from at least one probe molecule.

[00212] Aspect 42. The method of Aspect 41, wherein the signal is collected from at least two probe molecules, the at least two probe molecules differing from one another. This can be effected by having two different probe molecules bound to two different antibody-nucleic acid conjugates that are themselves bound to different proteins associated with a given EV. Different probe molecules can differ from one another in terms of nucleic acids, fluorophores, and the like.

[00213] Aspect 43. A system, comprising: an EV immobilized within a matrix, a fluorescence imager, and a compression element, wherein the fluorescence imager collects a signal indicative of a probe associated with a protein, the protein being associated with the EV, and wherein the compression element is optionally configured to compress the matrix to a height of less than about 6 pm, less than about 8 pm, less than about 10 pm, less than about 12 pm, or less than about 14 pm.

[00214] The compression element can be incorporated into a microscope; such an element can be a slide, a coverslip, or other element.

[00215] Aspect 44. A system, comprising: an EV immobilized within a matrix, an antibody associated with the EV, the antibody being conjugated to a pendant nucleic acid; a probe molecule bound to the pendant nucleic acid, the probe molecule comprising a probe nucleic acid and a fluorophore. Such a system can be considered a conjugate, in some instances. An example of such a conjugate is provided in FIG. 13.

[00216] References

[00217] Kalluri, R. & LeBleu, V. S. The biology, function, and biomedical applications of exosomes. Science (2020).

[00218] Morales, R. T. T. & Ko, J. Future of Digital Assays to Resolve Clinical Heterogeneity of Single Extracellular Vesicles. ACS nano (2022).

[00219] Mathieu, M., Martin-Jaular, L., Lavieu, G. & Thery, C. Specificities of secretion and uptake of exosomes and other extracellular vesicles for cell-to-cell communication. Nature cell biology (2019).

[00220] Niel, G. V., d’ Angelo, G. & Raposo, G. Shedding light on the cell biology of extracellular vesicles. Nature reviews Molecular cell ... (2018).

[00221] Ko, J., Wang, Y., Carlson, J. C. T. & Marquard. . ., A. Single extracellular vesicle protein analysis using immuno-droplet digital polymerase chain reaction amplification. Advanced ... (2020).

[00222] Hoshino, A., Kim, H. S., Bojmar, L., Gyan, K. E. & Cioffi. . ., M. Extracellular vesicle and particle biomarkers define multiple human cancers. Cell (2020).

[00223] Muraoka, S., DeLeo, A. M. & Sethi. . ., M. K. Proteomic and biological profiling of extracellular vesicles from Alzheimer’s disease human brain tissues. Alzheimer s & ... (2020).

[00224] Witwer, K. W. & Thery, C. Extracellular vesicles or exosomes? On primacy, precision, and popularity influencing a choice of nomenclature. Journal of extracellular vesicles (2019).

[00225] Chen, I. H., Xue, L., Hsu, C. C. & Paez. . ., J. S. P. Phosphoproteins in extracellular vesicles as candidate markers for breast cancer. Proceedings of the ... (2017).

[00226] Bernardi, S. & Balbi, C. Extracellular vesicles: From biomarkers to therapeutic tools. Biology (2020).

[00227] Ferguson, S., Yang, K. S., Zelga, P. & Liss..., A. S. Single-EV analysis (sEVA) of mutated proteins allows detection of stage 1 pancreatic cancer. Science ... (2022).

[00228] Lee, K. et al. Multiplexed profiling of single extracellular vesicles. ACS ...

(2018). [00229] Fraser, K., Jo, A., Giedt, J., Vinegoni, C. & Yang. . K. S. Characterization of single microvesicles in plasma from glioblastoma patients. Neuro ... (2019).

[00230] Kim, D., Woo, H. K., Lee, C., Min, Y. & Kumar. . . , S. EV-Ident: identifying tumor-specific extracellular vesicles by size fractionation and single-vesicle analysis. Analytical ... (2020).

[00231] Maia, J., Batista, S., Couto, N. & Gregorio. . ., A. C. Employing flow cytometry to extracellular vesicles sample microvolume analysis and quality control. Frontiers in Cell and ... (2020).

[00232] Lof, L., Ebai, T., Dubois, L., Wik, L. & Ronquist. . ., K. G. Detecting individual extracellular vesicles using a multicolor in situ proximity ligation assay with flow cytometric readout. Scientific reports (2016).

[00233] Yuyama, K., Sun, H. & Igarashi. . ., Y. Immuno-digital invasive cleavage assay for analyzing Alzheimer’s amyloid ss-bound extracellular vesicles. Alzheimer s ... (2022).

[00234] Wang, R., Zhao, X., Chen, X., Qiu, X. & Qing. . ., G. Rolling circular amplification (RCA)-assisted CRISPR/Cas9 cleavage (RACE) for highly specific detection of multiple extracellular vesicle MicroRNAs. Analytical ... (2019).

[00235] Yan, H., Wen, Y., Tian, Z., Hart, N. & Han. . ., S. A one-pot isothermal Casl2-based assay for the sensitive detection of microRNAs. Nature Biomedical ... (2023).

[00236] Basu, A. S. Digital assays part I: partitioning statistics and digital PCR. SLAS technology (2017).

[00237] Akama, K., Shirai, K. & Suzuki, S. Highly sensitive multiplex protein detection by droplet-free digital ELISA. Electronics and Communications ... (2019).

[00238] Rissin, D. M., Kan, C. W ., Campbell, T. G. & Howes..., S. C. Singlemolecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nature ... (2010).

[00239] Kiselinova, M., Pasternak, A. O. & Spiegelaere..., W. D. Comparison of droplet digital PCR and seminested real-time PCR for quantification of cell-associated HIV-1 RNA. PloSone (2014).

[00240] Ushio, R., Yamamoto, M., Nakashima, K. & Watanabe. . ., H. Digital PCR assay detection of circulating Mycobacterium tuberculosis DNA in pulmonary tuberculosis patient plasma. Tuberculosis (2016). [00241] Samiei, E., Tabrizian, M. & Hoorfar, M. A review of digital microfluidics as portable platforms for lab-on a-chip applications. Lab on a Chip (2016).

[00242] Yuan, J. & Sims, P. A. An automated microwell platform for large-scale single cell RNA-seq. Scientific reports (2016).

[00243] Manzoor, A. A. & Romita. . ., L. A review on microwell and microfluidic geometric array fabrication techniques and its potential applications in cellular studies. The Canadian Journal of ... (2021).

[00244] Wu, H. et al. Facile Method for Fabricating Microfluidic Chip Integrated with Microwell Arrays for Cell Trapping. Micromachines (2019).

[00245] He, J., Zhou, C., Xu, X., Zhou, Z. & Danoy. . ., M. Scalable Formation of Highly Viable and Functional Hepatocellular Carcinoma Spheroids in an Oxygen-Permeable Microwell Device for Anti-Tumor Drug Evaluation. Advanced ... (2022).

[00246] Lee, G. H., Lee, J. S., Oh, H. J. & Lee, S. H. Reproducible construction of surface tension-mediated honeycomb concave microwell arrays for engineering of 3D microtissues with minimal cell loss. PloS one (2016).

[00247] Zhang, F., Wang, S., Yin, L., Yang, Y. & Guan. . ., Y. Quantification of epidermal growth factor receptor expression level and binding kinetics on cell surfaces by surface plasmon resonance imaging. Analytical ... (2015).

[00248] Yang, K. S., Ciprani, D., O’Shea, A., Liss, A. S. & Yang..., R. Extracellular vesicle analysis allows for identification of invasive IPMN. Gastroenterology (2021).

[00249] Chen, G. et al. Exosomal PD-L1 contributes to immunosuppression and is associated with anti-PD-1 response. Nature (2018).

[00250] Yu, Z. L., Liu, J. Y. & Chen, G. Small extracellular vesicle PD-L1 in cancer: the knowns and unknowns. NPJ Precision Oncology (2022).

[00251] Garcia, L. M. C., Peterson, T. E. & Cepeda. . ., M. A. Isolation and analysis of plasma-derived exosomes in patients with glioma. Frontiers in ... (2019).

[00252] Li, C. et al. Clinical significance of PD-L1 expression in serum-derived exosomes in NSCLC patients. Journal of translational ... (2019).

[00253] Marcoux, G., Duchez, A. C., Cloutier, N. & Provost. . ., P. Revealing the diversity of extracellular vesicles using high-dimensional flow cytometry analyses. Scientific reports (2016). [00254] Bio, G. Multiplex Microarrays: The Next Generation ELISA. (2014).

[00255] Heo, S. J., Thakur, S., Chen, X., Loebel, C. & Xia..., B. Aberrant chromatin reorganization in cells from diseased fibrous connective tissue in response to altered chemomechanical cues. Nature biomedical ... (2023).

[00256] Thery, C., L. Zitvogel, and S. Amigorena, Exosomes: composition, biogenesis and function. Nature reviews immunology, 2002. 2(8): p. 569-579.

[00257] Kalluri, R. and V.S. LeBleu, The biology, function, and biomedical applications of exosomes. Science, 2020. 367(6478): p. eaau6977.

[00258] Shah, R., T. Patel, and J.E. Freedman, Circulating extracellular vesicles in human disease. New England Journal of Medicine, 2018. 379(10): p. 958-966.

[00259] Cheng, L., et al., Exosomes provide a protective and enriched source of miRNA for biomarker profiling compared to intracellular and cell-free blood. Journal of extracellular vesicles, 2014. 3(1): p. 23743.

[00260] Yim, K.H.W., et al., Assessing Extracellular Vesicles in Human Biofluids Using Flow-Based Analyzers. Advanced Healthcare Materials, 2023. 12(32): p. 2301706.

[00261] Allenson, K., et al., High prevalence of mutantKRAS in circulating exosome-derived DNA from early-stage pancreatic cancer patients. Annals of Oncology, 2017. 28(4): p. 741-747.

[00262] Wan, Y., et al., Nanoscale extracellular vesicle-derived DNA is superior to circulating cell-free DNA for mutation detection in early-stage non-small-cell lung cancer. Annals of Oncology 2018. 29(12): p. 2379-2383.

[00263] Ferguson, S. and R. Weissleder, Modeling EV kinetics for use in early cancer detection. Advanced biosystems, 2020. 4(12): p. 1900305.

[00264] Morales, R.-T.T. and J. Ko, Future of digital assays to resolve clinical heterogeneity of single extracellular vesicles. ACS nano, 2022. 16(8): p. 11619-11645.

[00265] Lee, K., et al., Multiplexed profiling of single extracellular vesicles. ACS nano, 2018. 12(1): p. 494-503.

[00266] Tian, Y., et al., Protein profiling and sizing of extracellular vesicles from colorectal cancer patients via flow cytometry. ACS nano, 2018. 12(1): p. 671-680.

[00267] Liu, C., et al., Single-exosome-counting immunoassays for cancer diagnostics. Nano letters, 2018. 18(7): p. 4226-4232. [00268] Daly, A.C., et al., Hydrogel microparticles for biomedical applications. Nature Reviews Materials, 2020. 5(1): p. 20-43.

[00269] Sorokin, N., et al., Kinetics of hybridization on surface oligonucleotide microchips: theory, experiment, and comparison with hybridization on gel-based microchips. Journal of biomolecular Structure and Dynamics 2006. 24(1): p. 57-66.

[00270] Krishnan, S., C.J. Weinman, and C.K. Ober, Advances in polymers for anti-biofouling surfaces. Journal of Materials Chemistry, 2008. 18(29): p. 3405-3413.

[00271] Jang, W., et al., Highly magnetized encoded hydrogel microparticles with enhanced rinsing capabilities for efficient microrna detection. Biomedicines, 2021. 9(7): p. 848.

[00272] Park, J., et al., High-Throughput Single-Cell, Single-Mitochondrial DNA Assay Using Hydrogel Droplet Microfluidics. Angewandte Chemie, 2024. 136(18): p. e202401544.

[00273] Lee, J.S., et al., Tissue beads: tissue-specific extracellular matrix microbeads to potentiate reprogrammed cell-based therapy. Advanced Functional Materials, 2019. 29(31): p. 1807803.

[00274] Jung, S., et al., Extensible multiplex real-time PCR of MicroRNA using Microparticles. Scientific reports, 2016. 6(1): p. 22975.

[00275] Kim, J., et al., Hydrogel-based hybridization chain reaction (HCR) for detection of urinary exosomal miRNAs as a diagnostic tool of prostate cancer. Biosensors and Bioelectronics, 2021. 192: p. 113504.

[00276] Loebel, C., et al., Shear-thinning and self-healing hydrogels as injectable therapeutics and for 3D-printing. Nature protocols, 2017. 12(8): p. 1521-1541.

[00277] Poldervaart, M.T., et al., 3D bioprinting of methacrylated hyaluronic acid (MeHA) hydrogel with intrinsic osteogenicity. PloS one, 2017. 12(6): p. e0177628.

[00278] Roh, Y.H., et al., Post-synthesis functionalized hydrogel microparticles for high performance microRNA detection. Analytica Chimica Acta, 2019. 1076: p. 110-117.

[00279] Roh, Y.H., et al., Vertically encoded tetragonal hydrogel microparticles for multiplexed detection of miRNAs associated with Alzheimer's disease. Analyst, 2016. 141(15): p. 4578-4586.

[00280] Al Sulaiman, D., N. Juthani, and P.S. Doyle, Quantitative and Multiplex Detection of Extracellular Vesicle-Derived MicroRNA via Rolling Circle Amplification within Encoded Hydrogel Microparticles. Advanced healthcare materials, 2022. 11(10): p. 2102332.

[00281] Mun, S. J., et al., Multiplex genotyping of SNPs in genomic DNA via hydrogel -based assay mediated with MutS and polyethylene glycol. Biosensors and Bioelectronics, 2023. 241 : p. 115670.

[00282] Zhang, F., et al., Quantification of epidermal growth factor receptor expression level and binding kinetics on cell surfaces by surface plasmon resonance imaging. Analytical chemistry, 2015. 87(19): p. 9960-9965.

[00283] Roh, Y.H., et al., Direct conjugation of streptavidin to encoded hydrogel microparticles for multiplex biomolecule detection with rapid probe-set modification. Polymers, 2020. 12(3): p. 546.

[00284] Reynolds, D.E., et al., Double Digital Assay for Single Extracellular Vesicle and Single Molecule Detection. Adv. Sci., 2023. 10: p. 2303619.

[00285] Ko, J., et al., Single extracellular vesicle protein analysis using immunodroplet digital polymerase chain reaction amplification. Advanced biosystems, 2020. 4(12): p. 1900307.

[00286] Spitzberg, J.D., et al., Multiplexed analysis of EV reveals specific biomarker composition with diagnostic impact. Nature communications, 2023. 14(1): p. 1239.

[00287] Horta, S., et al., Evaluation of immuno-rolling circle amplification for multiplex detection and profiling of antigen-specific antibody isotypes. Analytical chemistry, 2021. 93(15): p. 6169-6177.

[00288] Stirling, D.R., et al., CellProfiler 4: improvements in speed, utility and usability. BMC bioinformatics, 2021. 22: p. 1-11.

[00289] Han, C., et al., Single-vesicle imaging and co-localization analysis for tetraspanin profiling of individual extracellular vesicles. Journal of extracellular vesicles, 2021. 10(3): p. el2047.

[00290] Jeong, M.H., et al., Plasmon-Enhanced Single Extracellular Vesicle Analysis for Cholangiocarcinoma Diagnosis. Advanced Science, 2023. 10(8): p. 2205148.

[00291] Zubtsov, D., et al., Comparison of surface and hydrogel-based protein microchips. Analytical biochemistry, 2007. 368(2): p. 205-213. [00292] Dhand, A.P., et al., Simultaneous one-pot interpenetrating network formation to expand 3d processing capabilities. 2022. 34(28): p. 2202261.

Claims

What is Claimed:

1. A method, comprising: releasing at least one protein carried by an extracellular vesicle (EV) while the EV is in a container free of other EVs; and collecting a signal associated with the at least one protein.

2. The method of claim 1, further comprising preventing intermixing between the protein released from the EV and protein released from other EVs.

3. The method of any one of claims 1-2, further comprising immobilizing the at least one protein.

4. The method of claim 3, wherein the at least one protein is immobilized to a first antibody that binds to the at least one protein.

5. The method of claim 4, wherein the first antibody is associated with a capture microbead.

6. The method of claim 5, wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

7. The method of any one of claims 1-6, further comprising generating the signal.

8. The method of claim 7, wherein the signal is generated by tyramide signal amplification.

9. The method of any one of claims 1-8, further comprising classifying the EV based on the signal.

10. The method of any one of claims 1-9, further comprising classifying, based on the signal, a source of the EV.

11. The method of any one of claims 1-10, further comprising selecting a therapeutic treatment based on the signal.

12. A method, comprising: partitioning a plurality of EVs among a plurality of containers such that a given partitioned EV is in a container free of other EVs; releasing proteins from at least some of the partitioned EVs; collecting signals from the released proteins.

13. The method of claim 12, wherein the partitioning comprises contacting the plurality of EVs to a plurality of containers under conditions that give rise to partitioned EVs.

14. The method of claim 13, wherein the plurality of EVs is distributed among the plurality of containers according to a Poisson distribution with = 1.

15. The method of any one of claims 12-14, wherein a container has disposed therein (i) capture microbeads configured to bind to released proteins and (ii) spacer microbeads configured to be free of binding to released proteins.

16. The method of claim 15, wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

17. The method of any one of claims 12-15, further comprising generating the signal.

18. The method of claim 17, wherein the signal is generated by tyramide signal amplification.

19. The method of any one of claims 12-18, further comprising classifying a partitioned EV based on a collected signal.

20. The method of any one of claims 12-19, further comprising classifying a source of the partitioned EV based on a collected signal.

21. The method of any one of claims 12-20, further comprising selecting a therapeutic treatment based on a collected signal.

22. A system, comprising: a plurality of containers, the containers optionally comprised in a substrate; and a container having disposed therein (i) capture microbeads configured to bind to at least one protein released by an EV and (ii) spacer microbeads configured to be free of binding to the at least one protein, and wherein a capture microbead is separated from any other capture microbead by at least one spacer microbead.

23. The system of claim 22, wherein a container is a microwell, the microwell optionally being hexagonal in profile.

24. The system of claim 22, wherein a capture microbead comprises an antibody complementary to the at least one protein released by the EV.

25. The system of any one of claims 22-24, further comprising an imager configured to collect a signal associated with a protein released by an EV in a container.

26. A method, comprising: with an EV immobilized in a matrix, collecting a signal indicative of a labeled probe molecule being bound to a pendant nucleic acid associated with the EV.

27. The method of claim 26, wherein the matrix comprises a hydrogel.

28. The method of any one of claims 26-27, wherein the labeled probe molecule comprises a fluorophore conjugated to a probe nucleic acid.

29. The method of claim 28, wherein the probe nucleic acid is complementary to the pendant nucleic acid.

30. The method of claim 29, wherein the pendant nucleic acid is conjugated to an antibody.

31. The method of claim 30, wherein the antibody binds a protein associated with the EV.

32. The method of any one of claims 26-27, further comprising compressing the matrix.

33. The method of claim 32, further compressing the matrix by at least 10% in a direction.

34. The method of any one of claims 26-27, further comprising effecting rolling circle amplification on the pendant nucleic acid.

35. A method, comprising: immobilizing an EV within a matrix; contacting the EV with an antibody that binds to a protein associated with the EV, wherein the antibody is conjugated to a pendant nucleic acid; amplifying the pendant nucleic acid; contacting the pendant nucleic acid with a probe comprising a fluorophore conjugated to a probe nucleic acid, the probe nucleic acid being complementary to the pendant nucleic acid; collecting a signal indicative of the probe being bound to the protein.

36. The method of claim 35, wherein the matrix comprises a hydrogel.

37. The method of any one of claims 35-36, wherein the EV is immobilized through UV crosslinking.

38. The method of any one of claims 35-36, wherein the matrix is compressed, the matrix optionally being compressed to a height of less than about 6 pm, less than about 8 pm, less than about 10 pm, less than about 12 pm, or less than about 14 pm.

39. The method of any one of claims 35-36, wherein the matrix comprises at least about 1%, at least about 2%, at least about 2.5%, at least about 3%, at least about 4%, at least about 5%, at least about 6%, at least about 7%, at least about 8%, or at least about 9% methacrylated hyaluronic acid (MeHA).

40. The method of any one of claims 35-36, wherein from about 20 EVs to about 36 EVs are immobilized within a matrix.

41. The method of any one of claims 35-36, wherein the signal is collected from at least one probe molecule.

42. The method of claim 41, wherein the signal is collected from at least two probe molecules, the at least two probe molecules differing from one another.

43. A system, comprising: an EV immobilized within a matrix, a fluorescence imager, and a compression element, wherein the fluorescence imager collects a signal indicative of a probe associated with a protein, the protein being associated with the EV, and wherein the compression element is optionally configured to compress the matrix to a height of less than about 6 pm, less than about 8 pm, less than about 10 pm, less than about 12 pm, or less than about 14 pm.

44. A system, comprising: an EV immobilized within a matrix, an antibody associated with the EV, the antibody being conjugated to a pendant nucleic acid; a probe molecule bound to the pendant nucleic acid, the probe molecule comprising a probe nucleic acid and a fluorophore.