Methods for identifying and validating cell-type specific extracellular vesicle proteins

Abstract

Described herein are methods of identifying cell-type specific proteins that are transmembrane on or internal to extracellular vesicles (EVs).

Description

[0001] Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0002] METHODS FOR IDENTIFYING AND VALIDATING CELL-TYPE SPECIFIC EXTRACELLULAR VESICLE PROTEINS

[0003] CLAIM OF PRIORITY

[0004] This application claims the benefit of U.S. Provisional Application Serial No.

[0005] 63 / 763,834, filed on February 26, 2025. The entire contents of the foregoing are incorporated herein by reference.

[0006] TECHNICAL FIELD

[0007] Described herein are methods of identifying cell-type specific proteins that are transmembrane on or internal to extracellular vesicles (EVs).

[0008] BACKGROUND

[0009] Extracellular vesicles (EVs) are nanometer-scale, membrane-bound compartments that contain proteins, RNAs, and metabolites endogenous to their cell of origin1. As such, the content of EVs, isolated from biofluids, can serve as a molecular snapshot of the parent cell.

[0010] SUMMARY

[0011] Described herein are methods of identifying cell-type specific proteins that are transmembrane on or internal to extracellular vesicles (EVs). Identifying cell-type specific proteins associated with EVs allows non-invasive liquid biopsies for earlier disease diagnosis, for assessing treatment response, and for enhancing our mechanistic understanding of disease pathophysiology. However, the field of EV research has struggled to identify such markers due to an inability to differentiate true EV bound proteins vs nonspecifically bound contaminant proteins4,33. This has led to the use of targets that are predominantly secreted that are not in fact transmembrane on or internal to EVs, muddying the water for a liquid biopsy. In other cases, immunocapture of EVs has been performed with proteins entirely internal to EVs, which cannot represent a true immunocapture, as in the case of CNP3(described below).

[0012] Described herein are methods for identifying and validating EV associated proteins. These methods combine computational and experimental approaches to definitively identify EV association, allowing for the selection of transmembraneAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0013] proteins for EV immunocapture and for the selection of internal EV proteins for celltype of origin analysis and for characterization of the parent cell.

[0014] Provided herein are methods for identifying cell-type specific extracellular vesicle (EV)-associated proteins in a sample comprising a biofluid. The methods comprise: fractionating the sample into a plurality of fractions; selecting one or more EV fractions as comprising an EV-associated biomarker; identifying proteins present in the EV fractions; identifying each of the proteins identified as present in the EV fractions as transmembrane, internal, or external, and selecting proteins identified as transmembrane or external; and assigning each transmembrane or external protein to a cell type based on expression; thereby identifying cell-type specific extracellular vesicle (EV)-associated proteins.

[0015] In some embodiments, the biofluid comprises whole blood, plasma, serum, urine, sweat, saliva, lymph, cerebrospinal fluid (CSF), ascites, bronchoalveolar lavage fluid, pleural effusion, seminal fluid, sputum, nipple aspirate, post-operative seroma, or wound drainage.

[0016] In some embodiments, the EV-associated biomarker is tetraspanin (CD63), CD9, CD81, Annexin A2 (ANXA2), A4 (ANXA4), A5 (ANXA5), Alix (PDCD6IP), or Vacuolar protein sorting-associated protein VTA1 homolog (VTA1).

[0017] In some embodiments, the EV fractions are identified as comprising an EV-associated biomarker by a method comprising: dividing each fraction;

[0018] assaying each fraction for the presence of the EV-associated biomarker, optionally using an immunoassay; and selecting the one or more EV fractions that comprise an EV-associated biomarker.

[0019] In some embodiments, the EV fractions are identified as lacking non-EV markers by a method comprising, optionally wherein the non-EV markers comprise one or more of Fibronectin (FN1), Prothrombin (F2), Pigment epithelium-derived factor (PEDF, also known as SERPINF1), and Complement C3 (C3).

[0020] In some embodiments, identifying proteins present in the EV fractions comprises performing a multiplex enzyme-linked immunosorbent assay (ELISA).

[0021] In some embodiments, the methods further comprise: determining expression level of selected proteins present in EV fractions; determining expression level of the same selected proteins present in non-EV fractions; and assigning an EV association score for each protein by calculating a ratio of median expression level of the proteinAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0022] present in the EV fractions to the median expression level of the protein present in non-EV fractions.

[0023] In some embodiments, identifying each protein present in the EV fractions as transmembrane, internal, or external comprises using a database comprising information indicating subcellular localization of each protein, and / or using a machine learning algorithm to predict subcellular localization of each protein.

[0024] In some embodiments, assigning each transmembrane or external protein to a cell type comprises using one or more databases comprising information about expression of each protein in specific cell types.

[0025] In some embodiments, the methods further comprise performing a proteinase protection assay (PPA) to confirm transmembrane, internal, or external localization.

[0026] In some embodiments, the biofluid comprises cerebrospinal fluid (CSF), and the cell types comprise astrocytes, oligodendrocytes, microglia, neurons, and endothelial cells.

[0027] In some embodiments, fractionating the sample into a plurality of fractions comprises using size exclusion chromatography (SEC) or density gradient chromatography (DGC).

[0028] 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 to which this invention belongs. Methods and materials are described herein for use in the present invention; other, suitable methods and materials known in the art can also be used. The materials, methods, and examples are illustrative only and not intended to be limiting. All publications, patent applications, patents, sequences, database entries, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control.

[0029] Other features and advantages of the invention will be apparent from the following detailed description and figures, and from the claims.

[0030] DESCRIPTION OF DRAWINGS

[0031] Figure 1: CSF SEC fractionation as a measure of EV association.

[0032] la. Quantification of CSF fractions using a Simoa assay7for CD81. Four individual healthy CSF samples were fractionated using SEC and each fraction was analyzed by Simoa A Mann- Whitney U test performed comparing fractions 9 and 10Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0033] with fractions 7, 11, 12, and 13 in all samples combined showed fractions 9 and 10 are significantly greater than fractions 7, 11. 12. and 13 (p< 0.0005)

[0034] lb. Quantification of CSF fractions using the Olink assay for CD63. Four individual healthy samples were fractionated using SEC and each fraction was analyzed by the Olink HT panel. Mann-Whitney U test performed comparing fractions 9 and 10 with fractions 7, 11. 12, and 13 in all samples combined showed fractions 9 and 10 are significantly greater than fractions 7, 11, 12, and 13 (p<0.0005) lc. Heat maps showing normalized NPX values for each SEC fraction for four representative previously published EV contaminant and four EV-associated proteins in the Olink panel. Of note, the EV contaminant proteins F2, C3, FN1, and SERPINF1 (PEDF) all have increasingly high NPX values predominantly in the late free protein fractions. EV associated proteins ANXA2, ANXA4, ANXA5, and VTA1 show EV-associated fractionation patterns with high NPX values in fractions 9 and 10. Anxa5 and VTA1 also show NPX signals in later fractions 14 and 15 suggesting possibly soluble protein isoforms for these proteins.

[0035] ld. Percentage of Deep TMHMM predicted transmembrane, internal, and external targets quantified by Olink as having an EV fractionation pattern in CSF.

[0036] Figure 2: Cell-type specificity of proteins that show an EV-associated fractionation pattern. EV Association Score (EV-associated NPX signal in fractions 9 and 10 were greater than NPX signals in fractions 7, 11, 12, and 13) and calculated Tau Score of >0.75 for each identified transmembrane (red), internal (blue), and external (green) proteins that demonstrated an EV-associated fractionation pattern for a. Astrocytes b. Endothelial cells c. Microglia and d. Oligodendrocytes and e. Neurons.

[0037] Figure 3: Nanoparticle Tracking Analysis of E Vs in CSF SEC fractions 7-10 and 11-15. Particle size distribution was characterized in early (a) and late (b) SEC fractions. Red lines represent mean + / - standard error of the mean (SEM) for 5 measurements per sample.

[0038] Figure 4: Western blots of tetraspanins in CSF SEC fractions and human brain lysate (HBL). The SEC fractionation patterns of tetraspanins CD9 (a), CD63 (b), and CD81 (c) were characterized in CSF. Nonspecific signal was assessed by probing CSF SEC fractions and HBL with the same anti-mouse IgG secondaryAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0039] antibody used in Figure S2a-c but without primary antibody (d). All tetraspanins were detected in HBL and had the highest relative abundance in fractions 9 and 10.

[0040] Figure 5: Western blots of cell type-specific transmembrane proteins in CSF SEC fractions and HBL. The SEC fractionation patterns of select cell typespecific transmembrane proteins highlighted in Figure 2 were further assessed for EV association. These proteins include AQP1 (a), SLC16A1 (b), FCAR (c), CHRM3 (d). and TSPAN8 (e). and are transmembrane proteins specific to astrocytes, endothelial cells, microglia, neurons, and oligodendrocytes, respectively. Nonspecific signals were assessed by probing CSF SEC fractions and HBL with the corresponding secondary antibodies without the application of primary antibody used in Figure S3a-e. Anti-rabbit IgG (f), anti -mouse IgG (g), and anti-rat IgG (h) control for nonspecific secondary antibody binding in Figures 3a-c, 3d, and 3e, respectively. AQP1 and SLC16A1 were detected in HBL and had the highest relative abundance in fractions 9 and 10, confirming EV association. FCAR, CHRM3, and TSPAN8 detection was inconclusive due likely to variability in antibody quality and interreference of signal from the secondary antibody, as can be seen in the no primary antibody control.

[0041] Figure 6. Proteinase Protection Assay of control proteins. Albumin was used as a generic secreted protein while Alix was used as an internal EV protein.

[0042] Figure 7. Proteinase Protection Assay of S100A13, validating it as an internal target.

[0043] Figure 8. Proteinase Protection Assay of GFAP, demonstrating it is a contaminant protein in EV preparations.

[0044] Figure 9. Proteinase Protection Assay of CNPase, demonstrating it is entirely internal to EVs and therefore cannot be used for immunocapture.

[0045] DETAILED DESCRIPTION

[0046] A preponderance of EV research has focused on isolating EVs from specific cell types or tumors utilizing proteins annotated as transmembrane and enriched in the parent cell2. While this has led to some success, such as in the case of monitoring prostate cancer3, studies that have sought to capture brain-derived EVs have been hampered by methodological challenges. Specifically, proteins cited as transmembrane or internal to EVs have been shown to be predominantly cleaved and secreted4. It is, therefore, critical to validate methods to differentiate EV-associated proteins from those that are secreted and cleaved in biofluids.Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0047] Plasma and cerebrospinal fluid (CSF) EVs can be easily separated from soluble proteins by size exclusion chromatography (SEC) or density gradient chromatography (DGC)4. Nevertheless, analyzing the proteomic content of the EV and soluble protein fractions with a single biochemical technique can be difficult because the soluble protein fractions contain several orders of magnitude more protein than the EV fractions. Unbiased techniques like mass spectrometry are challenging because, in the EV fractions, lipoproteins can co-isolate and mask rare EV-associated proteins, while in the secreted protein fractions, abundant proteins like albumin create a similar problem5. Furthermore, the high levels of abundant proteins such as albumin in plasma, preclude the ability to use gel-based techniques such as Western blots. As a result, ELISAs have thus far been the best method of assessing EV fractionation pattern6. In previous work, we have utilized the ultrasensitive digital ELISA platform Simoa, to quantify canonical EV proteins (CD9, CD63, CD81, Alix), assess potential contaminants to EV preparations (Apolipoprotein B, albumin), and evaluate individual proteins as targets for cell-type specific enrichment4'6. Here, we sought to apply a large-scale unbiased method to generate a much-needed dataset and establish a bioinformatic approach to identify proteins that can be used for potential immunocapture of EVs secreted by a cell-type of interest, as well as cytosolic proteins to corroborate EV-brain-cell origin.

[0048] CSF directly surrounds the brain and the spinal cord, making CSF-derived EVs more likely to contain predominantly brain-specific markers compared to plasma and other biofluids7’8. Furthermore, CSF has an approximately 200-fold lower soluble protein content compared to plasma9. This lowers the chance of non-specific interactions compared to that for EVs isolated from plasma and other more complex matrices. Although estimates of the proportion of brain-derived EVs in CSF is highly limited by the lack of reliable markers, one study reported that approximately 16% of brain-specific proteins in CSF EVs were of neuronal origin while about 84% of them were of glial origin10. This makes CSF an ideal biofluid for brain-derived EV biomarker discovery analysis.

[0049] By utilizing the highly sensitive and specific multiplexed Olink platform on SEC -fractionated healthy CSF, we identified cell-type specific proteins that may be associated with EVs and can be used both for potential EV immunocapture and for the analysis of the luminal protein cargo of brain-derived EVs. Furthermore, weAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0050] demonstrate that 90% of predicted transmembrane proteins did not have a definable EV fractionation pattern, which we speculate is due to overwhelming signals from cleaved or secreted isoforms of these proteins. Such targets are likely not viable for use in EV immunocapture. Conversely, some targets identified as external were highly EV associated (e.g. EDIL3) and are likely bound tightly to the extravesicular surface making them potential immunocapture targets.

[0051] While previous research has explored EVs derived from brain tissues, cell-type specific media collected from induced pluripotent stem cells of various cell types, and CSF-derived EVs without any consideration of cell type specificity10,20’29, to our knowledge, this dataset provides the first unbiased proteomic profiling of EV association on a large scale, making it a valuable resource for future EV biomarker discovery. Additionally, with the growing importance of cell type-specific EVs in liquid biopsy for hard-to-biopsy organs (e.g., the brain), we have created a computational approach based on stringent criteria to discover potential cell typespecific brain-derived EV biomarkers. While many of the proteins we identified as EV associated have been described previously in the literature30-32, substantial additional work is required to assess cell origin for those proteins that meet the criteria displayed in Figure 2. Additional steps can be used to further validate these candidate proteins for each cell type, prioritizing those with the highest Tau and EV association scores. For example, validation can include using immunocapture with antibodies to a transmembrane or external protein and analysis of proposed internal targets with Simoa following a proteinase protection assay. The present methods provided a dataset that is a resource for identifying novel targets for brain-derived extracellular vesicles.

[0052] The present methods are performed on fractionated samples. Fractionated samples are samples that are divided, e.g., using size exclusion chromatography (SEC) or density gradient chromatography (DGC), into a number of fractions. For example, the samples can be divided into 10-30 fractions, 15-20 fractions, or more, e.g., 10-100 fractions. The methods are performed on fractions of the fractionated samples that comprise a large number of proteins, e.g., hundreds to thousands of different proteins, of unknown identity.

[0053] The present methods first include identification of proteins in fractionated samples. For example, a multiplexed ELISA can be run on fractionated biofluids ofAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0054] interest using any platform (herein we exemplified OLink17, however the same method can apply using Simoa, Luminex, SUPER or non-digital ELISAs) and markers identified as EV-associated based on the following criteria: A protein is considered to have a fractionation pattern typical of EV-associated proteins if the medians of fractions 9 and 10 is greater than the medians for fractions 7, 11, 12, and 13. While the exact fraction number can change based on the column run (herein we use a lOmL Sepharose 6B column), the point is to select the peak (which should be followed by a trough) mapped by EV-associated (e.g., tetraspanin (CD63)) proteins (which in this example are present in the highest abundance in fractions 9 and 10) while ignoring the free protein or secreted fractions6. The multiplexed ELISA can be configured for detection / identification of 2-20,000 different proteins, e.g., at least 2, 3, 4, 5, 6, 10, 12, 15, 20, 25, or 50, up to about 100, 1,000, 5,000, 10,000, or 20,000, or any range therebetween, e.g., 5-20,000, 50-20,000, 5-10,000, 50-10,000, or 5-5,000.

[0055] The present examples used CSF to look at brain-derived EVs, however the same process can be applied with any biofluid (e.g., whole blood, plasma, serum, urine, sweat, saliva, lymph, cerebrospinal fluid, ascites, bronchoalveolar lavage fluid, pleural effusion, seminal fluid, sputum, nipple aspirate, post-operative seroma or wound drainage).

[0056] Once proteins from the multiplexed ELISA are identified as EV associated, they are overlapped computationally with data on protein localization. In Example 1 below the DeepTMHMM dataset23was utilized, however the same analysis can be done by accessing information from other databases, which annotate proteins as transmembrane, cytosolic and secreted (e.g., the uniprot database). This allows identification of EV associated proteins that are internal or transmembrane for downstream analysis, and the selection of those identified proteins.

[0057] The identified protein list is then overlapped with cell-type specificity data using atlases of RNA or protein expression. In Example 1, we used the Brain RNAseq database14, however information from any single cell or bulk tissue atlas or database can be used (selected depending on the desired cell type), e.g., GTEx. human protein atlas, EMBL-EBI, or umGear.

[0058] The above steps can be used to identify cell-type specific EV associated proteins. To further confirm and validate EV association, a proteinase protection assay (PPA) can be performed4,33to identify false positive or erroneous hits. ForAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0059] example, if a protein is being characterized as internal it should completely recover after PPA, as in the case of the protein Alix, used here as a control internal protein (FIG. 6). In the present methods, recovery of 80-120% of the 'no treatment’ condition levels can be considered to be full recovery.

[0060] EXAMPLES

[0061] The invention is further described in the following examples, which do not limit the scope of the invention described in the claims.

[0062] Methods

[0063] The following methods were used in the Examples below.

[0064] Human sample preparation:

[0065] For the main experimental figures utilizing Olink and Simoa technology, one milliliter each of four healthy CSF samples (PrecisionMed) were thawed at room temperature and centrifuged at 2,000 g for 10 minutes. Subsequently, the supernatant from this first centrifugation was transferred to a 0.45-pm Coming Costar Spin-X filter (Sigma- Aldrich) and centrifuged again at 2,000 g for 10 minutes at room temperature. The flow through from this filtration was used for downstream experiments. For Supplementary Figures 1-3 (nanoparticle tracking analysis and Western blotting) one pooled lot of CSF (Innovative Research) was used to ensure enough material was available for all Western blots and nanoparticle tracking analysis without adding inter-individual variability7. The CSF was processed in the same way for this pooled lot as for the individual samples used in the main figures.

[0066] Size exclusion chromatography and fraction processing:

[0067] Sepharose CL-6B resin (GE Healthcare) was washed with an equal volume of PBS 3 times. For each wash, the resin was allowed to settle at 4°C overnight before the PBS was poured off and replaced. Following the washes, the resin was stored in an equal volume of PBS.

[0068] Econo-Pac Chromatography columns (Bio-Rad) were prepared immediately prior to fractionation. For each sample, washed resin was poured into a column to achieve a resin bed volume of 10.2 mL. A polyethylene bed support (Bio-Rad) was inserted into the top of the resin to compress to a bed volume of 10 mL. The packed resin was then washed with 20 mL of PBS. Immediately following the elution of the wash, 1 mL of each CSF sample was added to the respective column and fractionsAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0069] collected in 0.5 mL increments. When the 1 mL of CSF had flowed through, 0.5 mL of PBS was added to the column sequentially until fractions 1-15 were collected. Fractions 1-5 were discarded to avoid redundancy as EVs generally begin to elute in fractions 7 or 8 when using a lOmL Sepharose 6B column.

[0070] Each fraction 6-15 was transferred to a 10 kDa MWCO Ami con Ultra Centrifugal Filters (Sigma-Aldrich) and diluted to a total volume of 1.5 mL with PBS. These fractions were then centrifuged at 2,000 g at 4°C until all fractions were concentrated 15-fold. The concentrated fractions were brought to a volume of 97 pL with PBS. 76 pL was transferred to a 96 well plate supplied by Olink. Triton X-100 was added to a final concentration of 1% by volume, and the plate was stored at -80 °C. The remaining 21 pL of fraction volume was used to measure CD81 by Simoa.

[0071] Simoa CD81 sample analysis:

[0072] The Simoa analysis was performed according to the manufacturer’s instruction. Reagent preparation and assay parameters w ere followed as described previously in Norman & Ter-Ovanesyan et al. (2021)4. Abeam (anti-CD81 ab79559, clone M38) and Biolegend (anti-CD81 349502, clone 5A6) were used as capture and detector antibodies respectively. Human recombinant CD81 from Origene

[0073] (TP317508) was used in the calibration curve. Data analysis w as performed using GraphPad Prism version 10.1.1.

[0074] Nanoparticle tracking analysis:

[0075] Separate CSF fractions 7-10 and 11-15 were collected using size exclusion chromatography as described above. This 2mL volume w as condensed using a 10 kDa MWCO Amicon Ultra Centrifugal Filter (Sigma-Aldrich) to a volume of 500µL in PBS. Extracellular vesicle particle size and number were characterized using the NanoSight LM10 (Malvern Panalytical). 500µL of sample was injected and five, one-minute videos were captured at 24.98 fps with a detection threshold of 2 at a fixed temperature of 25 °C. Parameters were determined based on the manufacturer’s software manual and performed by the NTA 3.4 Build software v3.4.4.

[0076] Western sample analysis:

[0077] SEC was performed as above with 8mL of pooled CSF. Each mL w as loaded on its own column. Respective fractions were pooled and concentrated using 10 kDa MWCO Amicon Ultra Centrifugal Filters (Sigma-Aldrich). For fractions 6-12, one-sixteenth of the concentrated fractions were loaded per gel. For fractions 13-15,Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0078] protein input was normalized to fraction 12 to avoid overloading the gel. The fractions and human brain cerebellum whole tissue lysate (HBL) (Novus Biologicals) were denatured with 4x LDS and, for certain targets, reduced with DTT (see table below). Subsequently, CSF and HBL samples were heated at 70°C for 10 minutes, run at 150 V for 70 minutes on 4-12% Bolt Bis-Tris Plus gels (Thermo Fisher Scientific), and transferred to nitrocellulose membranes using the iBlot 3 Dry Blotting System (Thermo Fisher Scientific). The membranes were blocked for 30 minutes at 4°C and incubated with primary’ antibodies overnight. The next day, membranes were washed, incubated with secondary’ antibody (Bethyl Laboratories) for 1 hr at 4°C, and washed again. Nonspecific signals were assessed by probing CSF SEC fractions and HBL with the corresponding secondary antibodies (anti-mouse IgG, anti-rabbit IgG, or antirat IgG) without the application of primary' antibody. For primary antibody dilutions, secondary’ antibody dilutions, and membrane blocking, a PBS-T solution of 5% milk (weight by volume) with 1% Tween were used. All washes were performed with PBS-T (1% Tween) in cycles of three seven-minute washes (except SLC16A1, which was incubated in PBS-T six times per wash). Specifics on primary' antibodies used and dilutions can be found in the table below. After the final wash, blots were developed using the ProSignal Femto substrate kit (Genesee Scientific) and imaged with a Sapphire Biomolecular Imager (Azure Biosystems).

[0079] 1:25

[0080] Primary Primary Primary Primary Secondary Reducing Diluted HBL

[0081] Ta rget Antibody Antibody Antibody Antibody Antibody Conditions? Volume

[0082] Clone Vendor Species Dilution Dilution Loaded (ul)

[0083] M i II i pore

[0084] CD9 MM2 / 57 Mouse No 1 1:1000 1:2000 Sigma

[0085] BD CD63 H5C6 Mouse No 10 1:1000 1:2000 Biosciences

[0086] Thermo

[0087] CD81 M38 Fisher Mouse No 3 1:666 1:1000 Scientific

[0088] EPR11588

[0089] AQP1 Abeam Rabbit Yes 4 1:10000 1:1000 (B)

[0090] Cell Signaling

[0091] SLC16A1 E7A2K Rabbit Yes 1.75 1:1000 1:1000 Technology

[0092] FCAR EPR4622(2) Abeam Rabbit Yes 20 1:1000 1:1000 CHRM3 580011 R& D Systems Mouse Yes 20 1:1000 1:1000

[0093]

[0094] Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0095] TSPAN 458811 R& D Systems Rat Yes 20 1:1000 1:1000

[0096]

[0097] Olink sample analysis:

[0098] Samples were shipped on dry ice for analysis using the Olink HT platform, which measures 5416 unique proteins using highly multiplexed proximity extension assays. Pairs of antibodies with unique yet complimentary oligonucleotides called proximity' probes, each specific to a unique protein of interest, bind to their target antigens. After binding the target, the oligonucleotide probes encounter each other due to physical proximity' and hybridize resulting in the formation of an immunocomplex. The resulting hybridized proximity probes can be amplified by DNA polymerase creating a DNA amplicon that can be detected by quantitative PCR (qPCR) or next-generation sequencing (NGS) techniques2,11. Samples were run with a single replicate for each protein except for GBP1 and MAP2K1 which were run in Blocks 3.4 and 5 to check correlation between blocks.

[0099] The relative abundances of the amplicon, as measured by NGS, are then converted to normalized protein expression (NPX) values. The Olink panel includes plate, sample, and extension (ExtCtrl) controls. To ensure robustness, the NPX calculation accounts for variability in the different controls measured in the panel and includes a log 2 transformation of the data. The number of matched sequence reads (counts) generated by NGS is first normalized by the number of counts for the extension control of the sample, and then log2 transformed as follows:

[0100] / CountsfSample j Assay t

[0101] ExtNPXi = log 2

[0102]

[0103] \ C aunts (ExtCtr if)

[0104] where ExtNPX;, j is the NPX, normalized by the counts of the extension control specific to assay i measured in sample j. The median value of ExtNPX of the plate controls is then used to adjust for variability between plates, allowing comparison of relative protein abundances across the different plates11:

[0105] NPX[ j = ExtNPXi — median ExtNPXpiateii controi) control is the quality control measure collected from plate h, and NPXi is the reported NPX value for sample j, analyzed using assay i on plate h.

[0106] Further details on NPX value generation can also be found in the Olink website.Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0107] Data Analysis Methods:

[0108] The reported NPX values have an arbitrary unit that reflects the relative concentrations of the analyzed proteins in the sample of interest. All analysis was conducted in Python (version 3.11.5) using Visual Studio Code (Microsoft Corporation, Redmond, Washington). The HT panel includes 5420 proteins, including 5416 unique proteins and two assays processed in triplicate, measuring relative concentrations of MAP2K1 and GBP1, to ensure accuracy of data collection. Given that no calibration curve is included, all NPX values presented in fraction data are relative values only. The data points were all linearized, and two assays that were processed in triplicate were removed from dow nstream high throughput analysis.

[0109] The HT panel from Olink measures two negative controls for each assay. Olink recommends against calculating a limit of detection (LOD) with fewer than ten negative controls in a dataset, so we instead considered the fixed LODs made available by Olink. The fixed LOD calculation is based on 24-36 negative controls, ensuring a more robust calculation to minimize the higher variation among negative controls. This approach is consistent with the recommendations from Olink, which reports that values below LOD are unlikely to increase the risk of false positive discoveries and may be beneficial for biomarker discovery. They also highlight that filtering data based solely on LOD may remove meaningful signals, especially when a protein is well expressed in one group but undetectable in another. Therefore, excluding data points below LOD would prevent us from including potentially useful proteins in our analyses. The LOD data were not considered in downstream analyses.

[0110] Fractionation Analysis:

[0111] Four individual fractionated CSF samples (fractions 6-15) w ere submitted to Olink for analysis. Only fractions 7, 9, 10, 11, 12, and 13 were used to verify whether a protein exhibited the fractionation pattern typical for EV-associated proteins4. For each fraction of interest, the median NPX was calculated for each protein. A protein was considered to have the fractionation pattern ty pical of EV-associated proteins if the medians of fractions 9 and 10 were greater than the medians for fractions 7. 11.

[0112] 12, and 13.

[0113] Protein Localization:

[0114] The proteins in the Olink panel w ere computationally determined to be transmembrane, internal, or external using DeepTMHMM. DeepTMHMM is a deepAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0115] learning model-based algorithm that uses a hidden Markov model to predict subcellular localization of a protein in a cell. The model calculated a probability for each amino acid in each protein and returned the highest probability' domain for each amino acid, allowing the most likely' localization of the overall protein to be determined. Using this model, each amino acid was characterized as:

[0116] 1. Cytosolic

[0117] 2. Alpha transmembrane helix

[0118] 3. Beta transmembrane barrel

[0119] 4. Signaling peptide

[0120] 5. External to the cell and any secreted vesicles or exosomes

[0121] Information regarding signaling peptides were not considered, as they are largely cleaved from the protein when it enters the endoplasmic reticulum, and therefore would likely not be present in the epitope of the protein found in EVs12. We classified a protein as internal to the cell if all its amino acids were characterized as cytosolic and we classified a protein as external if all its amino acids were characterized as outside the cell or on secreted vesicles. Because proteins containing a transmembrane domain also contain domains found internal and external to the cell, a protein was classified as transmembrane if it contained one or more amino acids characterized as an alpha transmembrane helix or a beta transmembrane barrel.

[0122] Because of the budding mechanisms by which EVs are secreted13, it is largely assumed that proteins would have the same cytosolic, transmembrane, or extracellular domains in both EVs and the cell. Additional validation techniques can be used to confirm the localization of proteins relative to EVs, e.g., using SEC fractionation analysis as described previously.

[0123] EV-Associated Protein Identification:

[0124] This pipeline was used to identify proteins that may be associated with EVs. Proteins were labeled as internal to EVs if they met the fractionation criteria and were identified as internal using DeepTMHMM as described previously. The same criteria were followed to identify transmembrane and external proteins associated with EVs. This yielded a list which was further narrowed by selecting proteins considered to be cell-type specific based on the Tau score and BrainRNA-Seq dataset as described below. Each protein was assigned an “EV Association Score’", which was calculated as the ratio of the median NPX for the EV fractions (fractions 9 and 10), and theAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0125] median NPX for fractions 7, 11, 12, and 13. This value is shown on the y-axis of Figure 2.

[0126] Cell-Type Specificity:

[0127] The BrainRNA-Seq atlas reports fragments per kilobase per million mapped fragments (FPKM), collected via RNA sequencing14. The mean FPKM of each gene was used for mature astrocytes, neurons, oligodendrocytes, endothelial cells, and microglia. Fetal astrocytes were excluded from analysis. Tau specificity score is used to determine cell-type specificity' of genes, as it gives a numerical indication of the relative specificity of a gene across different cell types or tissues. Scores range between 0 and 1, where 0 indicates that a gene is ubiquitously expressed in all cell types, and 1 indicates that a gene is entirely expressed in a single cell type15. A gene was considered specific to a given cell type if it had a Tau specificity score of greater than 0.75 and if the mean FKPM was highest in the cell ty pe of interest relative to the other cell types. Tau specificity scores were calculated using the following formula15:

[0128] zr=i(i - ty)

[0129] T = - - - Tl — 1

[0130] Xi

[0131] xi = - max

[0132]

[0133] l<i<n

[0134] Xi = expression of the gene of interest in tissue i

[0135] n = number of tissues

[0136] The opposite is also true: by identifying proteins with a low Tau specificity score, <0.25, we selected genes that were ubiquitously expressed in all cell types. The genes were then mapped to proteins using data obtained from the UniProt website. This data is included in Table 3, below.

[0137] Brain Organ Specificity:

[0138] The GTEx database provides median gene-level expression transcripts per million (TPM) by tissue16. The tissues were grouped as described in Table 1 below:Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0139] TABLE 1

[0140] Group GTEx Portal Category

[0141] Brain Brain Amygdala

[0142] Brain Anterior cingulate cortex BA24 Brain Caudate basal ganglia Brain Cerebellar Hemisphere

[0143] Brain Cerebellum

[0144] Brain Cortex

[0145] Brain Frontal C ortex B A9

[0146] Brain Hippocampus

[0147] Brain Hypothalamus Brain_Nucleus_accumbens_basal_ganglia Brain Putamen basal ganglia Brain Spinal cord cerv ical c- 1 Brain Substantia nigra

[0148] Nerve_Tibial

[0149] Pituitary

[0150] Heart Heart_Atrial_Appendage

[0151] Heart Left Ventricle

[0152] Small Intestine Small lntestine T erminal ll eum Small_Intestine_Terminal_Ileum_Lymphode_Aggregate Small Intestine Terminal Ileum Mixed Cell Colon Colon Sigmoid

[0153] Colon_Transverse

[0154] Colon_Transverse_Mixed_Cell Colon_Transverse_Mucosa

[0155] Colon Transverse Muscularis

[0156] Liver Liver

[0157] Liver Hepatocyte

[0158] Liver_Mixed_Cell

[0159] Liver Portal Tract

[0160] Pancreas Pancreas

[0161] Pancreas_Acini

[0162] Pancreas_Islets

[0163] Pancreas Mixed Cell

[0164] Esophagus Esophagus_Gastroesophageal_Junction

[0165] Es ophagus_Mucos a

[0166] Esophagus Muscularis

[0167] Stomach Stomach

[0168] Stomach_Mixed_Cell

[0169] Stomach_Mucosa

[0170] Stomach Muscularis

[0171] Kidney Kidney Cortex

[0172] Kidney Medulla

[0173] Adipose Adipose_Subcutaneous

[0174] Adipose Visceral Omentum

[0175] Artery Artery _Aorta

[0176] Artery _C oronary

[0177]

[0178] Artery TibialAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0179] Group GTEx Portal Category

[0180] Skin Skin_Not_Sun_Exposed_Suprapubic

[0181] Skin Sun Exposed Lower leg

[0182] Muscle Muscle_Skeletal

[0183] Cervix Cervix Ectocervix

[0184] Cervix Endocervix

[0185] Lung Lung

[0186] Spleen Spleen

[0187] Testis Testis

[0188] Breast Breast Mammary Tissue

[0189] Ovary7Ovary'

[0190] Prostate Prostate

[0191] Thyroid Thyroid

[0192] Bladder Bladder

[0193] Uterus Uterus

[0194] Vagina Vagina

[0195] Cell Culture Cells_Cultured_fibroblasts

[0196] Cells EBV-transformed lymphocytes

[0197] Fallopian tube Fallopian Tube

[0198] Minor salivary gland Minor Salivary Gland

[0199] Adrenal gland Adrenal Gland

[0200]

[0201] Whole blood Whole Blood

[0202] The median TPM of each organ group was used to calculate the tissue specificity of each gene. Tau specificity score was used to determine the organ specificity of each gene, as it gives a numerical indication of the relative specificity of a gene across different organ groups. Scores ranged between 0 and 1, where 0 indicates that a gene is ubiquitously expressed in all tissue, and 1 indicates that a gene is entirely specific to a single tissue type15. This data was not considered in quantifying cell-type specificity as shown in Figure 2.

[0203] Example 1. Identification of Markers for Brain-Derived Extracellular Vesicles in Cerebrospinal Fluid

[0204] We used a highly multiplexed proximity extension assay platform from Olink to analyze thousands of proteins from microliters of biofluid with high specificity17. To assess EVs coming from the brain, we fractionated CSF from healthy individuals using SEC to separate proteins that peak in the early EV fractions from those that peak in the late secreted protein fractions18,19.

[0205] To define our EV fractions, adhering to MISEV 201818and 202319guidelines, we analyzed 20% of each fraction using our previously validated Simoa assay forAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0206] CD81 to demonstrate that EVs predominantly eluted in fractions 9 and 10 (Figure la)4-6. The remaining 80% of each fraction was analyzed using the Olink HT platform, which quantifies 5416 unique proteins. We analyzed the fractionation pattern of CD63 (tetraspanin) using data from the Olink assay and demonstrated a peak in signal in fractions 9 and 10 (Figure lb). We performed nanoparticle tracking analysis to show that EV-sized particle counts are increased in fractions 7-10 (Figures 3A-B). Next, we performed Western blots of CD9, CD63 and CD81 on fractionated CSF and demonstrated peak signals in fractions 9 and 10 for all three tetraspanins. Of note, in the Olink data, CD63 had a second later peak, which was not observed in Western blotting indicating this peak may be caused by nonspecific binding in the setting of high protein abundance in the later soluble protein fractions. Finally, in agreement with the literature16’20'22and MISEV 201818, we also report several previously identified generic EV markers, Annexins A2 (ANXA2), A4 (ANXA4), and A5 (ANXA5), and Vacuolar protein sorting-associated protein VTA1 homolog (VTA1),) and non-EV contaminant markers Fibronectin (FN1), Prothrombin (F2), Pigment epithelium-denved factor (PEDF. also known as SERP1NF1), and Complement C3 (C3) (Figure 1c) included in our Olink pipeline.

[0207] To identify targets that could be effective for EV immunocapture or for the analysis of EV cargo, we selected all proteins where the median normalized protein expression (NPX) value across CSF samples was greater in both fractions 9 and 10 compared to fractions 7, 11, 12, and 13. Because many proteins can be found as both EV -bound and soluble isoforms, we did not consider relative protein abundance in fractions 14 and 15 in our criteria, but rather selected proteins where a definable EV fractionation pattern could be seen. The signal from EV-associated proteins begins to peak from fraction 8 and reaches its highest point in fractions 9 and 10. With minimal to no signal observed in fractions 6 and 7, we treat these two fractions as internal controls. However, due to proportional signals from fractions 6 and 7 based on both CD81 Simoa (Figure la) and CD63 Olink (Figure lb) assays, we used fraction 7, rather than the combination of fractions 6 and 7. in our EV fractionation pattern selection criteria. Next, we utilized the DeepTMHMM deep learning model to differentiate cytosolic, transmembrane, and external proteins. Running this model on each protein analyzed by the Olink platform, we categorized them into 953 predicted transmembrane, 3522 predicted cytosolic, and 941 predicted external proteins23. WeAttorney Docket No. 29618-0536WO1 / BWH2024-543

[0208] demonstrated that 80% of predicted cytosolic proteins, 10% of transmembrane proteins, and 9% of external proteins have a definable EV fractionation pattern (Figure Id).

[0209] The HT panel from Olink measures two negative controls for each assay. Olink recommends against calculating a lower limit of detection (LOD) with fewer than ten negative controls in a dataset, so we instead considered the fixed LODs made available by Olink. The fixed LOD calculation is based on 24-36 negative controls, ensuring a more robust calculation to minimize the higher variation among negative controls. When thresholding our data using the LODs, we observed a significant loss of many targets. In total, without considering LOD, our analysis pipeline identified 57 unique transmembrane and internal proteins associated with EVs. However, when we thresholded using the fixed LOD, this number dropped to three proteins across the five cell types. This is consistent with the recommendations from Olink, which reports that values below LOD are unlikely to increase the risk of false positive discoveries and may risk eliminating informative biomarkers. They also highlight that filtering data based solely on LOD may remove meaningful signals, especially when a protein is well expressed in one group but undetectable in another. For example, Aquaporin 1 (AQP1), a transmembrane protein specific to astrocytes is eliminated from consideration when thresholding due to all fractions except fraction 9 being below LOD. However, we independently validated its fractionation pattern via Western blot (Figures 5A-H) and received results similar to those reported through Olink.

[0210] Therefore, excluding data points below LOD could prevent us from including potentially useful proteins in our analyses.

[0211] Our primary interest in using this dataset was to identify proteins that can be used to isolate or define an EV’s cell of origin. Therefore, we overlaid the Olink data with the BrainRNA-Seq atlas and selected proteins that were enriched in a specific brain cell-type - as defined by having a Tau specificity score >0.7514-24, calculated using the mean astrocyte, oligodendrocyte, microglia, neuron, and endothelial cell expression levels. Thus, we identified candidate transmembrane and external proteins that can potentially be used in CSF to isolate cell-type specific brain-derived EVs as well as candidate cytosolic proteins that can be analyzed as internal EV cargo to confirm cell-type specificity following immunocapture (Figure 2, Table 2).Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0212] TABLE 2

[0213] unipr ev_ brain_ body_ sym cell_ ev_ associati sym name localization tau_ tau_ bol

[0214] ids on_score type score score Q.166 1.735784 OCL endothel 0.8697 0.8525 25 218 N occludin TMhelix ial 01 04

[0215] CAC calcium voltage-gated

[0216] Q139 1.201866 NA1 channel subunit alphal 0.8638 0.8341 36 136 C C TMhelix neuron 07 39 P240 1.129486 FCA 0.9861 0.9909 71 693 R Fc alpha receptor TMhelix microglia 63 36 P299 5.978439 AQP aquaporin 1 (Colton astrocyt 0.9274 0.7696 72 475 1 blood group) TMhelix e 83 97 P539 2.877614 SLC1 solute carrier family 16 endothel 0.8533 0.7649 85 411 6A1 member 1 TMhelix ial 2 11 P190 1.556105 TSP oligoden 0.8242 0.9167 75 597 AN8 tetraspanin 8 TMhelix drocyte 94 99 P203 1.532309 CHR cholinergic receptor 0.9650 0.8651 09 73 M3 muscarinic 3 TMhelix neuron 51 36 P484 1.156062 SOX SRY-box transcription astrocyt 0.8747 0.9490 31 061 2 factor 2 internal e 54 74 Q9N 13.74318 FBX 0.7626 0.7090 WN3 282 034 F-box protein 34 internal neuron 77 06 Q6Z 1.184658 BEN BEN domain containing 0.9427 0.9489 U67 422 D4 4 internal neuron 8 79 nuclear protein 1,

[0217] 0603 1.782987 NUP transcriptional astrocyt 0.8261 0.6655 56 655 R1 regulator internal e 45 4 Q9N 11.11325 0.8745 0.7859 YY3 878 PLK2 polo like kinase 2 internal neuron 95 5

[0218] CIO

[0219] Q96 2.450563 orf9 chromosome 10 open oligoden 0.9435 0.9812 M02 008 0 reading frame 90 internal drocyte 66 14 Q2V2 1.525290 FHO formin homology 2 0.8396 0.8469 M9 501 D3 domain containing 3 internal neuron 53 98

[0220] G protein-coupled

[0221] GPR receptor associated

[0222] Q6PI 1.269690 ASP sorting protein family 0.8803 0.8231 77 036 3 member 3 internal neuron 24 64 Q96C 1.178444 SYTL astrocyt 0.8776 0.8900 24 862 4 synaptotagmin like 4 internal e 81 86 P143 1.321139 HCL hematopoietic cell0.9581 0.9071 17 734 SI specific Lyn substrate 1 internal microglia 37 18 Q6ZV 22.48144 FGD FYVE, RhoGEF and PH astrocyt 0.8485 0.6041 73 479 6 domain containing 6 internal e 06 69

[0223] ArfGAP with SH3

[0224] Q8T 1.232706 ASA domain, ankyrin repeat astrocyt 0.8702 0.7107 DY4 085 P3 and PH domain 3 internal e 04 21 A1YP 1.182340 ZBT zinc finger and BTB astrocyt 0.7943 0.7849

[0225]

[0226] RO 105 B7C domain containing 7C internal e 29 2Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0227] unipr ev_ brain_ body_ sym cell_ ev_ associati sym name localization tau_ tau_ bol

[0228] ids on_score type score score SHR

[0229] Q8TF 1.431518 00 shroom family member astrocyt 0.8994 0.7097 72 945 M3 3 internal e 32 66 0960 1.103646 PAK p21 (RAC1) activated endothel 0.8216 0.5644 13 001 4 kinase 4 internal ial 99 21

[0230] NADH:ubiquinone

[0231] NDU oxidoreductase

[0232] Q9P0 2.292007 FAF complex assembly 0.7860 0.3744 32 402 4 factor 4 internal neuron 67 51

[0233] ZC3

[0234] A2A2 23.20279 H12 zinc finger CCCH-type 0.7865 0.8953 88 057 D containing 12D internal microglia 97 9 receptor interacting

[0235] Q9Y5 2.042535 RIPK serine / threonine kinase 0.7691 0.6750 72 103 3 3 internal microglia 44 99 Q9Y2 1.178818 PDE 0.7927 0.8893 33 949 10A phosphodiesterase 10A internal neuron 38 21

[0236] HO

[0237] Q9N 1.216009 MER homer scaffold protein 0.7977 0.7823 SB8 636 2 2 internal neuron 75 12 Q150 2.055802 R3H R3H domain containing 0.8168 0.6932 32 831 DM1 1 internal neuron 2 01 Q8TE 1.177743 PAR par-3 family cell astrocyt 0.7667 0.6378 WO 526 D3 polarity regulator internal e 21 73 Q7Z6 1.453282 NRA NOTCH regulated 0.7537 0.7834 K4 528 RP ankyrin repeat protein internal microglia 65 34 Q9U 20.18835 PHF 0.7562 0.9718 PV7 004 24 PHD finger protein 24 internal neuron 26 35 Q128 4.012006 AKA A-kinase anchoring 0.9284 0.6634 02 763 P13 protein 13 internal microglia 89 05 Q96 1.458298 C5or chromosome 5 open 0.7578 0.7901 MH7 515 f34 reading frame 34 internal neuron 63 51 P203 2.327496 RAB RAB3B, member RAS 0.9762 0.9231 37 478 3B oncogene family internal neuron 78 3 Q166 1.194689 TBR T-box brain 0.9518 0.9959 50 183 1 transcription factor 1 internal neuron 42 32

[0238] S10

[0239] Q995 1.728405 0A1 S100 calcium binding astrocyt 0.8697 0.7247 84 754 3 protein A13 internal e 81 31

[0240] TO translocase of outer

[0241] Q153 1.364204 MM mitochondrial 0.7589 0.4359 88 27 20 membrane 20 internal neuron 41 48 0608 3.516015 GAS 0.7854 0.6993 61 977 7 growth arrest specific 7 internal neuron 46 5 Q5SY 1.111169 CLV 0.9491 0.9355

[0242]

[0243] Cl 384 S2 clavesin 2 internal neuron 3 76Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0244] unipr ev_ brain_ body_ sym cell_ ev_ associati sym name localization tau_ tau_ bol

[0245] ids on_score type score score Q9BT 1.595474 sox SRY-box transcription endothel 0.9566 0.8773 81 684 7 factor 7 internal ial 22 85 Q9N 1.188933 FGF fibroblast growth factor 0.9122 0.8779 P95 464 20 20 internal microglia 6 57 Q9BY 1.235875 STK3 serine / threonine kinase astrocyt 0.8966 0.9059 T3 802 3 33 internal e 52 28 P300 1.461420 PRD astrocyt 0.8176 0.5009 41 198 X6 peroxiredoxin 6 internal e 48 86

[0246] GPR G protein-coupled

[0247] Q5JY 1.477610 ASP receptor associated 0.9245 0.8047 77 422 1 sorting protein 1 internal neuron 13 47 Q9BY 1.484093 PAR par-6 family cell 0.8068 0.7268 G5 671 D6B polarity regulator beta internal neuron 13 72 pleckstrin homology

[0248] Q86S 1.596671 PHL like domain family B endothel 0.9130 0.6526 Q0 219 DB2 member 2 internal ial 83 66 P198 12.30762 NFK nuclear factor kappa B 0.8982

[0249] 38 169 Bl subunit 1 internal microglia 81 0.6606 Q9P2 1.368908 JCA junctional cadherin 5 endothel 0.9693 0.6754 66 513 D associated internal ial 98 04 P541 2.095868 oligoden 0.8806 0.9277 32 058 BLM BLM RecQ like helicase internal drocyte 75 15

[0250] MIS

[0251] Q.6P0 1.733139 18B 0.9409 0.7462 NO 147 Pl MIS18 binding protein 1 internal microglia 83 91 A6NI 1.763575 CCD coiled-coil domain endothel 0.8006 0.6940 79 086 C69 containing 69 internal ial 38 8 0152 1.466266 PFD 0.7762 0.6191 12 557 N6 prefoldin subunit 6 internal neuron 62 18 P577 1.267463 CAB calcium binding protein 0.7696 0.9100 96 685 P4 4 internal microglia 98 07 Q9N 1.294861 TSH teashirt zinc finger 0.9098 0.7738 RE2 489 Z2 homeobox 2 internal neuron 15 23 P052 3.332351 FGF fibroblast growth factor oligoden 0.8712 0.9606 30 228 1 1 internal drocyte 15 62 Q9H 1.215634 BH betaine--homocysteine astrocyt 0.9260 0.9217 2M3 058 MT2 S-methyltransferase 2 internal e 06 57

[0252] TNF

[0253] Q6P5 1.306130 AIP8 TNF alpha induced 0.9296 0.9211

[0254]

[0255] 89 282 L2 protein 8 like 2 internal microglia 18 48

[0256] Finally we identified a set of proteins that demonstrated a clear EV fractionation pattern but were not specific to a given cell type as defined by a Tau score <0.25 (Table 3). These latter proteins can be used to normalize total EV quantity.Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0257] TABLE 3

[0258] uniprot- symbol name tau_score ev_association_ ids score Localization: internal

[0259] ankyrin repeat and sterile alpha motif

[0260] Q6ZW76 ANKS3 domain containing 3 0.098157 2.041889 Q96CA5 BIRC7 baculoviral IAP repeat containing 7 0.035714 1.546482 Q9NY30 BTG4 BTG anti-proliferation factor 4 0.17901 1.929462 Q5SXH7 PLEKHS1 pleckstrin homology domain containing SI 0.035714 1.463884

[0261] CCDC17

[0262] P0C7W6 2 coiled-coil domain containing 172 0.035714 3.053878

[0263] SLC19A4 solute carrier family 19 member 4,

[0264] Q53S99 P pseudogene 0.035714 1.344501 C9J069 AJM1 apical junction component 1 homolog 0.035714 1.352159 Q8TC20 CAGE1 cancer antigen 1 0.035714 1.543021 P30307 CDC25C cell division cycle 25C 0.061511 1.322758 Q92817 EVPL envoplakin 0.035714 1.353568 A8MZ36 EVPLL envoplakin like 0.035714 1.598589

[0265] FAM186 family with sequence similarity 186

[0266] A6NE01 A member A 0.079129 1.337131 P41235 HNF4A hepatocyte nuclear factor 4 alpha 0.035714 1.467094

[0267] KIAA161

[0268] Q5VZ46 4 KIAA1614 0.078056 1.776628

[0269] KLHDC7

[0270] Q96G42 B kelch domain containing 7B 0.035714 1.236712

[0271] LCK proto-oncogene, Src family tyrosine

[0272] P06239 LCK kinase 0.035714 2.433822

[0273] CCDC19

[0274] Q8NCU1 7 coiled-coil domain containing 197 0.035714 1.442384 Q6P5Q4 LMOD2 leiomodin 2 0.155654 2.152395 P43355 MAGEA1 MAGE family member Al 0.035714 1.536187 Q9UBF1 MAGEC2 MAGE family member C2 0.035714 2.952618 015049 N4BP3 NEDD4 binding protein 3 0.051329 1.413891 000221 NFKBIE NFKB inhibitor epsilon 0.060877 1.763129 Q7Z3Z3 PIWIL3 piwi like RNA-mediated gene silencing 3 0.035714 1.478879 P30613 PKLR pyruvate kinase L / R 0.035714 1.351532

[0275] PPP1R14 protein phosphatase 1 regulatory inhibitor

[0276] Q9NXH3 D subunit 14D 0.165717 1.205489 Q9H0F5 RNF38 ring finger protein 38 0.035714 1.474461 Q9NXZ1 SAG El sarcoma antigen 1 0.035714 1.745016

[0277] SERPINB

[0278] P36952 5 serpin family B member 5 0.035714 1.367625

[0279] SH3 and multiple ankyrin repeat domains

[0280] Q9Y566 SHANK1 1 0.237016 1.076024

[0281] SOWAH sosondowah ankyrin repeat domain family

[0282] A6NEL2 B member B 0.087988 1.165361 Q8TDD2 SP7 Sp7 transcription factor 0.099082 1.161318

[0283]

[0284] P22528 SPRR1B small proline rich protein IB 0.035714 1.600335Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0285] uniprot_ symbol name tau_score ev_association_ ids score Q8IYJ3 SYTL1 synaptotagmin like 1 0.035714 8.293181

[0286] TBC1D2

[0287] Q2M2D7 8 TBC1 domain family member 28 0.035714 1.449551 Q9NZQ9 TMOD4 tropomodulin 4 0.035714 1.178077

[0288] Localization: transmembrane

[0289] Q8NFR9 IL17RE interleukin 17 receptor E 0.170594 1.536118 Q8N6P7 IL22RA1 interleukin 22 receptor subunit alpha 1 0.035714 1.356529

[0290] MARVEL

[0291] Q96A59 D3 MARVEL domain containing 3 0.083366 3.033138 Q86UW2 SLC51B SLC51 subunit beta 0.035714 1.086684

[0292]

[0293] Q92911 SLC5A5 solute carrier family 5 member 5 0.035714 14.2521

[0294] Example 2. Validation of Markers Identified in Brain-Derived Extracellular Vesicles in Cerebrospinal Fluid

[0295] The study described in Example 1 provided a list of cell-type specific EV associated proteins. To confirm and validate EV association and localization (e.g. internal vs transmembrane vs external), a proteinase protection assay (PPA) can be performed4,33. This allows erroneous hits to be discarded. If a protein is being characterized as internal it should completely recover after PPA, as in the case of the protein Alix, used here as a control internal protein (FIG. 6). We considered levels equivalent to 80-120% of the ‘no treatment’ condition to be full recovery.

[0296] Thus, using a PPA optimized to fully digest external proteins like CD81 and Albumin but to preserve internal proteins like Alix, we interrogated the targets from Example 1. For instance, we identified S100A13 as a predicted internal protein derived from astrocytes in CSF. As expected, when we performed PPA, S100A13 completely recovered (FIG. 7). Thus, we concluded that it is an internal EV associated protein coming from astrocytes in the CSF.

[0297] This is in stark contradistinction to GFAP, a commonly used protein which is said to define astrocyte origin that was in fact fully digested in CSF and plasma, indicating it is an external contaminant to EVs35. See FIG. 8. This was unexpected and important, because DeepTMHMM as used in Example 1 predicted that GFAP is completely internal, but the data suggested that in biofluids it is found free and not associated with EVs and. importantly, cannot be found internal to EVs even using high sensitivity assays. This was confirmed with this second step of PPA.Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0298] When dealing with transmembrane proteins one must determine if the epitopes of the ELISA used are for a fully internal or external epitope. If internal, 100% recovery is expected by PPA, as shown above for S100A13 and Alix. If either antibody binds an external epitope, complete digestion is to be expected (as seen above for Albumin and GFAP, FIGs. 6 and 8).

[0299] Such validation is key, because prior publications have claimed to conduct immunocapture with proteins that we found to be entirely internal, like CNPase (FIG.

[0300] 9), which is clearly not possible3. Therefore, this additional validation is an important part of the present methods and should be implemented across EV research, specifically when diagnostics and liquid biopsy procedures are involved.

[0301] Therefore, the multistep methods described herein allow for the discovery and validation of true cell-type specific proteins internal to or transmembrane on EVs for liquid biopsy of any cell ty pe, which previously has been done incorrectly leading to results based on nonspecific binding as opposed to identifying changes in an organ of interest.

[0302] References

[0303] 1 Raposo, G. & Stoorvogel, W. Extracellular vesicles: exosomes, microvesicles, and friends. J Cell Biol 200, 373-383, doi: 10.1083 / jcb.201211138 (2013).

[0304] 2 Shami-Shah, A., Norman, M. & Walt, D. R. Ultrasensitive Protein Detection Technologies for Extracellular Vesicle Measurements. Mol Cell Proteomics 22, 100557, doi:10.1016 / j.mcpro.2023.100557 (2023).

[0305] 3 Ramirez-Garrastacho, M. et al. Extracellular vesicles as a source of prostate cancer biomarkers in liquid biopsies: a decade of research. Br J Cancer 126, 331-350, doi:10.1038 / s41416-021-01610-8 (2022).

[0306] 4 Norman, M. et al. LI CAM is not associated with extracellular vesicles in human cerebrospinal fluid or plasma. Nat Methods 18, 631-634,

[0307] doi:10.1038 / s41592-021-01174-8 (2021).

[0308] 5 Ter-Ovanesyan, D. et al. Improved isolation of extracellular vesicles by removal of both free proteins and lipoproteins. Elife 12, doi: 10.7554 / eLife.86394 (2023).Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0309] 6 Ter-Ovanesyan, D. et al. Framework for rapid comparison of extracellular vesicle isolation methods. Elife 10. doi: 10.7554 / eLife.70725 (2021).

[0310] 7 Hladky, S. B. & Barrand, M. A. Mechanisms of fluid movement into, through and out of the brain: evaluation of the evidence. Fluids Barriers CNS 11, 26, doi:10.1186 / 2045-8118-11-26 (2014).

[0311] 8 Shetgaonkar, G. G. et al. Exosomes as cell-derivative carriers in the diagnosis and treatment of central nervous system diseases. DrugDeliv Transl Res 12, 1047-1079, doi: 10.1007 / s 13346-021- 1026-0 (2022).

[0312] 9 Fogh, J. R., Jacobsen, A. M., Nguyen, T., Rand, K. D. & Olsen, L. R. Investigating surrogate cerebrospinal fluid matrix compositions for use in quantitative LC-MS analysis of therapeutic antibodies in the cerebrospinal fluid. Anal Bioanal Chem 412, 1653-1661, doi: 10.1007 / s00216-020-02403-3 (2020).

[0313] 10 Muraoka, S. et al. Proteomic Profding of Extracellular Vesicles Derived from Cerebrospinal Fluid of Alzheimer's Disease Patients: A Pilot Study. Cells 9. doi:10.3390 / cells9091959 (2020).

[0314] 11 Wik, L. et al. Proximity Extension Assay in Combination with Next-Generation Sequencing for High-throughput Proteome-wide Analysis. Mol Cell Proteomics 20, 100168, doi: 10.1016 / j.mcpro.2021.100168 (2021).

[0315] 12 Liaci, A. M. & Forster, F. Take Me Home, Protein Roads: Structural Insights into Signal Peptide Interactions during ER Translocation. IntJMol Sci 22. doi:10.3390 / ijms222111871 (2021).

[0316] 13 Teng, F. & Fussenegger, M. Shedding Light on Extracellular Vesicle Biogenesis and Bioengineering. Adv Sci (Weinh) 8, 2003505,

[0317] doi: 10.1002 / advs.202003505 (2020).

[0318] 14 Zhang, Y. et al. Purification and Characterization of Progenitor and Mature Human Astrocytes Reveals Transcriptional and Functional Differences with Mouse. Neuron 89, 37-53, doi:10.1016 / j.neuron.2015.11.013 (2016).

[0319] 15 Kryuchkova-Mostacci, N. & Robinson-Rechavi, M. A benchmark of gene expression tissue-specificity metrics. Brief Bioinform 18, 205-214, doi:10.1093 / bib / bbw008 (2017).

[0320] 16 Kowal, J. et al. Proteomic comparison defines novel markers to characterize heterogeneous populations of extracellular vesicle subtypes. Proc Natl Acad Sci USA 113, E968-977, doi: 10.1073 / pnas.1521230113 (2016).Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0321] 17 PEA - a high-multiplex immunoassay technology with qPCR or NGS readout, olink.com / content / uploads / 2021 / 09 / olink-white-paper-pea-a-high-multiplex-immunoassay-technology-with-qpcr-or-ngs-readout-vEO.pdf (2020).

[0322] 18 Thery, C. et al. Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines. J Extracell Vesicles 7, 1535750, doi: 10.1080 / 20013078.2018.1535750 (2018).

[0323] 19 Welsh, J. A. et al. Minimal information for studies of extracellular vesicles (MISEV2023): From basic to advanced approaches. J Extracell Vesicles 13, e12404, doi:10.1002 / jev2.12404 (2024).

[0324] 20 You, Y. et al. Human neural cell type-specific extracellular vesicle proteome defines disease-related molecules associated with activated astrocytes in Alzheimer's disease brain. J Extracell Vesicles 11, e12183, doi:10.1002 / jev2.12183 (2022).

[0325] 21 Jeppesen, D. K. et al. Reassessment of Exosome Composition. Cell 177, 428-445 e418, doi: 10.1016 / j.cell.2019.02.029 (2019).

[0326] 22 Hallal, S., Tuzesi, A., Grau, G. E., Buckland, M. E. & Alexander, K. L. Understanding the extracellular vesicle surface for clinical molecular biology. J Extracell Vesicles 11, e12260, doi:10.1002 / jev2.12260 (2022).

[0327] 23 Hallgren, J. et al. DeepTMHMM predicts alpha and beta transmembrane proteins using deep neural networks. bioRxiv, 2022.2004.2008.487609, doi:10.1101 / 2022.04.08.487609 (2022).

[0328] 24 Camargo, A., Vasconcelos, A., Fiamenghi, M., Pereira, G. & Carazzolle, M. (Research Square, 2020).

[0329] 25 Aebersold, R. et al. How many human proteoforms are there? Nat Chem Biol 14, 206-214, doi: 10.1038 / nchembio.2576 (2018).

[0330] 26 Consortium, G. T. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science 369, 1318-1330, doi:10.1126 / science.aazl776 (2020).

[0331] 27 Tang, K., Page, J. S. & Smith, R. D. Charge competition and the linear dynamic range of detection in electrospray ionization mass spectrometry. J Am Soc Mass Spectrom 15, 1416-1423, doi:10.1016 / j.jasms.2004.04.034 (2004).Attorney Docket No. 29618-0536WO1 / BWH2024-543

[0332] 28 A, G. M. et al. Mass resolution and mass accuracy: how much is enough? Mass Spectrom (Tokyo) 2, S0009, doi:10.5702 / massspectrometry. S0009 (2013).

[0333] 29 Muraoka, S. et al. Proteomic and biological profiling of extracellular vesicles from Alzheimer's disease human brain tissues. Alzheimers Dement 16, 896-907, doi: 10.1002 / alz.12089 (2020).

[0334] 30 Oshikawa-Hori, S.. Yokota-Ikeda, N., Sonoda. H.. Sasaki. Y. & Ikeda, M. Reduced urinary release of AQP1- and AQP2-bearing extracellular vesicles in patients with advanced chronic kidney disease. Physiol Rep 9, el 5005,

[0335] doi:10.14814 / phy2.15005 (2021).

[0336] 31 Hoshino, A. et al. Extracellular Vesicle and Particle Biomarkers Define Multiple Human Cancers. Cell 182, 1044-1061 el018,

[0337] doi:10.1016 / j.cell.2020.07.009 (2020).

[0338] 32 Gupta, K. et al. Necroptosis is associated with Rab27-independent expulsion of extracellular vesicles containing RIPK3 and MLKL. J Extracell Vesicles 11, e12261, doi:10.1002 / jev2.12261 (2022).

[0339] 33 Gilboa, T. et al. Assessment of extracellular vesicle protein cargo as neurodegenerative disease biomarkers. bioRxiv, 2023.2011.2002.565137, doi:10.1101 / 2023.11.02.565137 (2023).

[0340] 34 Yu, Z. et al. Reduced oligodendrocyte exosome secretion in multiple system atrophy involves SNARE dysfunction. Brain 143, 1780-1797,

[0341] doi:10.1093 / brain / awaa110 (2020).

[0342] 35 Goetzl, E. J. et al. Cargo proteins of plasma astrocyte-derived exosomes in Alzheimer's disease. FASEB J 30, 3853-3859,

[0343] doi: 10.1096 / fj.201600756R (2016).

[0344] OTHER EMBODIMENTS

[0345] It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

Attorney Docket No. 29618-0536WO1 / BWH2024-543WHAT IS CLAIMED IS:

1. A method of identifying cell-type specific extracellular vesicle (EV)-associated proteins in a sample comprising a biofluid, the method comprising: fractionating the sample into a plurality of fractions;selecting one or more EV fractions as comprising an EV-associated biomarker; identifying proteins present in the EV fractions;identifying each of the proteins identified as present in the EV fractions as transmembrane, internal, or external, and selecting proteins identified as transmembrane or external; andassigning each transmembrane or external protein to a cell type based on expression;thereby identifying cell-type specific extracellular vesicle (EV)-associated proteins.

2. The method of claim 1, wherein the biofluid comprises whole blood, plasma, serum, urine, sweat, saliva, lymph, cerebrospinal fluid (CSF), ascites, bronchoalveolar lavage fluid, pleural effusion, seminal fluid, sputum, nipple aspirate, post-operative seroma, or wound drainage.

3. The method of claim 1, wherein the EV-associated biomarker is tetraspanin (CD63), CD9, CD81, Annexin A2 (ANXA2), A4 (ANXA4), A5 (ANXA5), Alix (PDCD6IP), or Vacuolar protein sorting-associated protein VTA1 homolog (VTA1).

4. The method of claim 1, wherein the EV fractions are identified as comprising an EV-associated biomarker by a method comprising:dividing each fraction;assaying each fraction for the presence of the EV-associated biomarker, optionally using an immunoassay; andselecting the one or more EV fractions that comprise an EV-associated biomarker.

5. The method of claim 1, wherein the EV fractions are identified as lacking non-EV markers by a method comprising, optionally wherein the non-EV markers comprise one or more of Fibronectin (FN1), Prothrombin (F2), PigmentAttorney Docket No. 29618-0536WO1 / BWH2024-543epithelium-derived factor (PEDF, also known as SERPINF1), and Complement C3 (C3).

6. The method of claim 1, wherein identifying proteins present in the EV fractions comprises performing a multiplex enzyme-linked immunosorbent assay (ELISA).

7. The method of claims 1-6, further comprising:determining expression level of selected proteins present in EV fractions; determining expression level of the same selected proteins present in non-EV fractions; andassigning an EV association score for each protein by calculating a ratio of median expression level of the protein present in the EV fractions to the median expression level of the protein present in non-EV fractions.

8. The method of claim 1, wherein identifying each protein present in the EV fractions as transmembrane, internal, or external comprises using a database comprising information indicating subcellular localization of each protein, and / or using a machine learning algorithm to predict subcellular localization of each protein.

9. The method of claim 1, wherein assigning each transmembrane or external protein to a cell type comprises using one or more databases comprising information about expression of each protein in specific cell types.

10. The method of claim 1, further comprising performing a proteinase protection assay (PPA) to confirm transmembrane, internal, or external localization.

11. The method of claim 1, wherein the biofluid comprises cerebrospinal fluid (CSF), and the cell types comprise astrocytes, oligodendrocytes, microglia, neurons, and endothelial cells.

12. The method of claim 1, wherein fractionating the sample into a plurality of fractions comprises using size exclusion chromatography (SEC) or density gradient chromatography (DGC).

Images