Methods of detection and uses of PD-l1

Abstract

The present disclosure relates to methods for identifying and treating type 1 diabetes. More specifically, the present disclosure relates to identifying and measuring programmed cell death ligand 1 (PD-L1) in extracellular vesicles emanating from beta cells and treating patients for type diabetes if the amount of PD-L1 is above a threshold level.

Description

METHODS OF DETECTION AND USES OF PD-L1CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Patent Application No. 63 / 554002 filed on February 15, 2024, the disclosure of which is expressly incorporated herein.STATEMENT OF GOVERNMENT SUPPORT

[0002] This invention was made with government support under DK133881 awarded by the National Institutes of Health. The government has certain rights in the invention.TECHNICAL FIELD

[0003] The present disclosure relates to methods for identifying and treating type 1 diabetes. More specifically, the present disclosure relates to identifying and measuring programmed cell death ligand 1 (PD-L1) in extracellular vesicles emanating from beta cells and treating patients for type diabetes if the amount of PD-L1 is above a threshold level.BACKGROUND

[0004] The global prevalence of type 1 diabetes is increasing, with current modeling predicting a prevalence of 13.5-17.4 million individuals by 2040. Mechanistic understanding of factors impacting disease progression is critical to understanding heterogeneity in type 1 diabetes progression as well as in the development of disease-modifying therapies. An increasing body of evidence demonstrates a critical role of intrinsic beta cell signaling in the potentiation of immune cell pathogenicity in type 1 diabetes. One such pathway involves immune checkpoints, such as programmed cell death-ligand 1 (PD-L1), a normal component of the immune system that provides negative feedback on immune cells. PD-L1 which is found on the surface of cells potentially targeted by the immune system binds to its receptor, programmed cell death protein 1 (PD-1), on immune cells to inhibit activity. Immune checkpoint inhibitors, including PD-L1 and PD-1 inhibitors, are used in certain cancers to increase antitumor immune responses against cancer cells. Beta cells also express PD-L1 , and while checkpoint inhibitor treatment frequently results in autoimmune complications including type 1 diabetes, PD-L1 upregulation prevents diabetes in mouse models.

[0005] Extracellular vesicles (EVs) are membrane-bound nanoparticles that can be generated by an endosomal sorting complex required for transport-dependent multi-vesicularbody formation or by outward budding of the plasma membrane. Cargo carried in EVs depends on the healthy and diseased state of individuals. EVs play a critical role in cell-cell communications. EVs have multiple immunoregulatory functions and cytokine-induced stress results in the release of EV cargo with inflammatory properties that promote B-cell autoimmunity in type 1 diabetes. Moreover, changes in small EV cargo in the context of beta cell proinflammatory cytokine exposure activates the CXCL10 / CXCR3 pathway that increases beta cell dysfunction while lymphocyte-derived exosomes containing the microRNAs miR- 142- 3p, miR-142-5p and miR-155 promote beta cell death.

[0006] In cancer cells, PD-L1 expression occurs on EVs to allow tumor cells to evade the immune system. However, the presence of PD-L1 in beta cell EVs has not been used to identify treatments for type 1 diabetes. The present disclosure is directed to identifying the upregulation of PD-L1 in islet beta cell EVs (EV PD-L1) during islet beta cell inflammation before initiating treatment for type 1 diabetes.

[0007] The disclosure is directed to the identification of EV PD-L1 as an additional mechanism supporting the effects of the PD-L1 / PD-1 axis in type 1 diabetes both at the level of the islet and in circulation. These findings identify a plausible framework for EV PD-L1 binding to PD-1, allowing a mechanistic basis for immune modulation, and suggesting that EV PD-L1 could serve as a predictive biomarker during development of type 1 diabetes.SUMMARY

[0008] The present disclosure includes one or more of the features recited in the appended claims and / or the following features which, alone or in any combination, may comprise patentable subject matter.

[0009] According to a first aspect of the present disclosure, a method for treating type 1 diabetes in a subject comprises measuring a first amount of PD-L1 is circulation, comparing the first amount of PD-L1 to a baseline value of PD-L1, and treating the subject by providing a therapy for type 1 diabetes if the first amount of PD-L1 is higher than the baseline amount of PD-L1.

[0010] In some embodiments, the first amount of PD-L1 is associated with extracellular vesicles formed from beta cell. In some embodiments, the PD-L1 is located on the surface of the extracellular vesicle. In some embodiments, the extracellular vesicles are exosomes. In some embodiments, the extracellular vesicles are microvesicles or apoptotic bodies. In some embodiments, the extracellular vesicles are CD81+ extracellular vesicles. In someembodiments, number of extracellular vesicles comprising the amount of PD-L1 in circulation is not substantially different than number of extracellular vesicles comprising the baseline value of PD-L1. In some embodiments, interferon exposure increases the amount of PD-L1 in circulation to be more than the baseline value. In some embodiments, the interferon is IFN-a. In some embodiments, the interferon is IFN-y. In some embodiments, the amount of PD-L1 in circulation suppresses proliferation and cytotoxicity of CD8 cells. In some embodiments, the PD-L1 in circulation can bind to PD-1. In some embodiments, the method further comprises determining a stage of type 1 diabetes of the subject based on the amount of PD-L1 in circulation. In some embodiments, a change in the baseline value is indicative of an early stage of type 1 diabetes. In some embodiments, the amount of PD-L1 in circulation is measured in plasma of the subject.

[0011] According to a second aspect of the present disclosure, A method of preventing autoimmune beta cell destruction comprises delivering extracellular vesicles loaded with PD- L1 to beta cells. In some embodiments, the method comprises delivering extracellular vesicles loaded with PD-L1 to immune cells.

[0012] BRIEF DESCRIPTION OF DRAWINGS

[0013] The following description accompanies the drawing(s), all given by way of nonlimiting examples that may be useful to understand how the described method and composition may be embodied.

[0014] FIG. 1 illustrates that PD-L1 is present in beta cell EVs. (1A) Confocal microscopy of PD-L1 (red) and tetraspanin-associated surface markers EV-associated tetraspanins CD63 (yellow), and CD9 (green) in INS-1 cells. (IB) TEM analysis of INS-1 cell-derived EVs. TEM images display representative data from INS-1 EVs. Scale bars, 500 nm. (1C) Immunoblot of INS1 cells and EVs for PD-L1, EV markers (CD63 and CD9) and calreticulin (reflects cellular contamination). n=3. (ID) PD-L1 colocalizes with beta cell EV-associated proteins in EndoC- [3H1 cells. Confocal microscopy of PD-L1 (green) and tetraspanin-associated surface markers CD63 (red), and CD81 (yellow) in EndoC-[3Hl cells.

[0015] FIG. 2 illustrates that IFN-a or IFN-y increases beta cell EV PD-L1 cargo. (2A-2D) Immunoblot of (2A-2B) INS1 and (2C-2D) EndoC-PHl cells and EVs + / -24-h IFN-a for PD- Ll, EV markers (CD63 and CD9) and calreticulin. (2B, 2D) Quantification of EV PD-L1 protein expression was performed using densitometry and normalized to total protein content (Ponceau S stain) from (2B) INS-1 and (2D) EndoC-PHl EVs. (2E) Immunoblot of EndoC-PHl cells andEVs + / -24-h IL1-P, IFN-y or IL1-J3 and IFN-y for PD-L1, CD63 and calreticulin. (2F) Quantification of EV PD-L1 protein expression was performed using densitometry and normalized to total protein content (Ponceau S stain). (2G) PD-L1 in human islet EVs + / -24-h IFN-a as measured using ELISA. (2H) Human islet EV fold change in EV PD-L1 + / -24-h IFN- a. Human islet soluble PD-L1 + / -24-h IFN-a as measured using ELISA. (21) PD-L1 in human islet EVs + / -24-h IFN-y as measured using ELISA (2J-2K) NTA showing EV size distribution of (2H) INS1, (21) EndoC-PHI and (2J) human islet EVs + / -24-h IFN-a. Blue bars / lines / circles indicated vehicle control, red bar / lines / circles indicate IFN-a treated samples, and green bar / lines / circles indicate IFN-y treated samples. N=3 for INS-1 and EndoC-PHI cells; n=3-6 for human islets. Results are displayed as mean ± SEM *,p value<.05; ***,p valueC.OOl, ****, p value <.0001.

[0016] FIG 3 illustrates that PD-L1 protein is present on the beta cell EV surface and can bind PD-1. (3A) PD-L1 EV surface expression as quantified by flow cytometry at baseline or after exposure to IFNa (red) or IFN-y (green). INS- 1 EVs + / - 24-h IFN-a were captured with biotin labelled CD9 or CD63 beads and Exo-FITC levels represented as dot plot from unstained or bead pulldown. Histograms represent changes in mean fluorescence intensity of PD-L1 in b (grey), vehicle control (blue), IFN-a treated (red), or IFN-y treated (green). (3B) The geometric mean of the mean fluorescence intensity (MFI) was quantified at baseline and after exposure to IFNa or IFN-y. Data represented as mean ± SEM. Blue circles indicated vehicle control, red circles indicate IFN-a treated samples, and green circles indicate IFN-y treated samples. (3C) Serial dilution of EndoC-PHI EV (blue) and PD-L1 standard (grey) as a positive control. (3D) Immunoblot of HEK293 cells with or without PD-L1 overexpression. P actin is the loading control. (3E) Serial dilution of EVs from HEK293 wildtype (grey) and HEK293 cells overexpressing PD-L1 (purple). Dose response curves determined by normalized HTRF ratio using non-linear regression analysis. Data represented as mean ± SEM; n=3-4; *, p value<.05; ***, p value<.001; ****, p value<.0001. WT: wildtype.

[0017] FIG. 4 illustrates that IFN-a or IFN-y induce PD-L1 expression within EV tetraspanin-associated EV subpopulations. Capture spot image of (4A-4B) EndoC-PHI and (4C-4D) total fluorescent particle count of PD-L1 in tetraspanin-associated populations of EndoC-PHI EVs (4C) + / -24-h IFN-a or (4D) + / -24-h IFN-y. Capture spot image of (4E-4F) human islets and (4G-4H) Total fluorescent particle count of PD-L1 in tetraspanin-associated populations of human islet EVs (g) + / -24-h IFN-a or (h) + / -24-h IFN-y. Data represented as mean ± SEM; n=3-4 for EndoC-[3Hl EVs and n=4-5 for human islets EVs; *, p value<.05. Bluecircles indicated vehicle control, red circles indicate IFN-a treated samples, and green circles indicate IFN-y treated samples.

[0018] FIG. 5 illustrates that EV PD-L1 inhibits proliferation, cytokine production and cytotoxicity of murine CD8 T cells in vitro. (5A-5E) Line graph representing the proportion of cells with indicated treatments positive for (5A) CD8+ T cells (SB) CD8+ T cells with diluted CTV to assess proliferation (SC) CD8+CD69+ expression as a measure of CD8+ T cell activation, (5D) CD8+CD25+ expression as a measure of CD8+ T cell activation and (5E) CD8+CD44+ expression as a measure of CD8+ T cell activation. (5F) Representative histogram of CTV-labelled NOD CD8 T cells with or without NIT-1 EV treatment at 72 h. (5G) Bar graphs of proportion of cells with diluted CTV dye. (5H-5M) Representative contour plots and corresponding bar graphs showing proportion of cells from NOD CD8 T cells examined for the surface expression of activation markers: (5H-5I) CD69, (5J-5K) CD25 and (5L-5M) CD44. (5N) Granzyme secretion and (50) IFN-y secretion from the NOD splenocytes with or without EV treatment as measured by ELISA. Data represented as mean± SEM; n=3; *, p value<.05; ****, p value<.0001; ns, nonsignificant.

[0019] FIG. 6 illustrates that EV PD-L1 inhibits proliferation, cytokine production and cytotoxicity of murine CD8 T cells in vitro. (6A) Immunoblot of NIT- 1 cells with or without PD-L1 overexpression. P actin is the loading control. (6B) PD-L1 levels from EVs derived from NIT-1 cells with or without PD-L1 overexpression (PD-L1 OE). (6C-6D) Flow cytometric analysis of CD8 T cells isolated from mouse splenocytes. Gating strategy and representative plots showing (6C) proliferation assessed by Cell Trace Violet (CTV) and (6D) activation status of CD8 T cells; determined by CD69, CD25 and CD44 surface expression.

[0020] FIG. 7 illustrates that Plasma EV PD-L1 levels are higher in islet autoantibody positive individuals compared to controls and plasma EV PD-L1 correlates with circulating C- peptide levels in individuals with islet autoantibody positivity or recent-onset type 1 diabetes. (7A-7D) In plasma from children with recent-onset type 1 diabetes vs. non-diabetic controls: (7A) Total EV particle number (7B) EV size distribution (7C) EV PD-L1 (7D) Soluble PD-L1 and (7E) plasma C-peptide levels. (7F) Linear regression analysis of plasma EV-PD-L1 and plasma C-peptide levels from children with type 1 diabetes (red, T1D) and non-diabetic (blue, non-TID controls). (7G-7J) In plasma from individuals positive for islet autoantibodies (Aab+) vs autoantibody negative (Aab-) controls: (7G) EV PD-L1 (7H) Soluble PD-L1 levels (71) plasma C-peptide levels, (7J) Linear regression analysis of plasma EV-PD-L1 and plasma C- peptide levels from islet autoantibody positive (green, Aab+) and autoantibody negative (blue,Aab- controls). Data shown as mean ± SEM, a-c, e-g; n=26 / group, d; n=17 / group; g-k; n=16 / Aab+, n=20 / Aab- controls; *, p value<.05; ****, p value<.0001.DETAILED DESCRIPTION

[0021] Certain exemplary embodiments will now be described to provide an overall understanding of the principles of the structure, function, manufacture, and use of the devices and methods disclosed herein. One or more examples of these embodiments are illustrated in the accompanying drawings. Those skilled in the art will understand that the devices and methods specifically described herein and illustrated in the accompanying drawings are nonlimiting exemplary embodiments and that the scope of the present disclosure is defined solely by the claims. The features illustrated or described in connection with one exemplary embodiment may be combined with the features of other embodiments. Such modifications and variations are intended to be included within the scope of the present disclosure. Accordingly, aspects and features of every embodiment may not be described with respect to each embodiment, but those aspects and features are applicable to the various embodiments unless statements or understandings are to the contrary.

[0022] As used herein, the term “patient” or “user” refers to any subject including mammals and humans. The patient may have a disease or suspected of having a disease and as such is being treated with a drug. In some instances, the patient is a mammal, such as a human, a premature neonate, neonate, infant, juvenile, adolescent, or adult thereof. In some instances, the term “patient,” as used herein, refers to a human (e.g., a man, a woman, or a child). In some instances, the term “patient,” as used herein, refers to laboratory animal of an animal model study. The patient or subject may be of any age, sex, or combination thereof.

[0023] The term “treating” refers to administering a therapy in an amount, manner, or mode effective (e.g., a therapeutic effect) to improve a condition, symptom, disorder, or parameter associated with a disorder, or a likelihood thereof.

[0024] The terms “essentially” or “substantially” as used herein mean to a great or significant extent, but not completely.

[0025] The term “about” as used herein refers to any values, including both integers and fractional components that are within a variation of up to ±10% of the value modified by the term “about.”

[0026] The present disclosure is directed to systems and methods for treating type 1diabetes. In some embodiments, PD-L1 may be present in the extracellular vesicles (EV) of pancreatic beta cells (beta cells). In some embodiments, the amount of PD-L1 present in the extracellular vesicles of beta cells may increase when the beta cells are in a pro-inflammatory type 1 diabetes microenvironment. In some embodiments, patients of subjects may be treated for type 1 diabetes based on the identification of PD-L1 in extracellular vesicles emanating from beta cells. In some embodiments, the PD-L1 in extracellular vesicles emanating from beta cells may protect the beta cells by inactivating immune cells. In some embodiments, the amount of PD-L1 in the extracellular vesicles emanating from beta cells may be increased during type 1 diabetes due to beta cell inflammation. In some embodiments, a patient or subject may be treated for type 1 diabetes based on an increase in the amount of PD-L1 in extracellular vesicles emanating from beta cells compared to amounts found in healthy individuals. In some embodiments, the amount of PD-L1 in extracellular vesicles emanating from beta cells is used as an indicator for treating subjects at risk for type 1 diabetes.

[0027] In some embodiments, PD-L1 is present on the surface of small EVs secreted by beta cells. In some embodiments, PD-L1 is upregulated in EVs in response to IFN-a or IFN-y in beta cells and / or primary human islets. In some embodiments, PD-L1 present in EVs may bind to PD-1 receptors. In some embodiments, EV PD-L1 may have the capacity to bind to PD-1 and significantly impair the proliferation, activation, and / or cytotoxicity of activated CD8 T cells. In some embodiments, circulating EV PD-L1 may increase in individuals with islet autoantibody positivity compared to non-diabetic controls. In some embodiments, circulating EV PD-L1 may increase in individuals with single positive autoantibodies compared to nondiabetic controls. In some embodiments, circulating PD-L1 in EVs may be positively associated with C-peptide in humans with type 1 diabetes. In some embodiments, levels of circulating EV PD-L1 may positively correlate with residual beta cell function in autoantibody positive individuals as well as in individuals with recent onset type 1 diabetes. In some embodiments, circulating PD-L1 in EVs may play a protective role in humans.

[0028] IFNs upregulate PD-L1 expression in human beta cells but not by other cytokines. IFN-a plays a prominent role in mediating host viral responses and in the autoimmune response during early type 1 diabetes development. A type 1 IFN-a gene signature is present in individuals during early type 1 diabetes development, with elevated plasma IFN-a. In the human islet microenvironment, IFN-a exposure can increase beta cell PD-L1 expression via JAK / STAT-IRF1 signaling. In some embodiments, IFN-a exposure increases total cell PD-L1 expression. In some embodiments, IFN-a or IFN-y exposure increases PD-L1 shuttling into EVcargo in multiple EV subpopulations.

[0029] In some embodiments, beta cell PD-L1 on the EV surface can directly bind to PD-1 , with functional studies demonstrating the immunomodulatory capacity of beta cell EV PD-L1. In some embodiments, beta cells may employ EV PD-L1 as an additional mechanism to modulate autoreactive immune responses in type 1 diabetes serving as a paracrine signal to amplify immune checkpoint signaling by the beta cell. In some embodiments, there may be a positive relationship between circulating EV PD-L1 with residual C-peptide in individuals with autoantibody positivity or clinical type 1 diabetes.

[0030] EV membrane staining with an antibody recognizing the extracellular domain of PD- L1 confirmed the presence of PD-L1 on the EV surface. In some embodiments, IFN-a upregulates the levels of PD-L1 localized to the EV membrane (EV PD-L1). EV PD-L1 can directly bind to PD-1, suggesting that EV PD-L1 could feasibly bind to PD-1 on immune cells to exert a protective effect in type 1 diabetes autoimmunity. This is also illustrated in FIG. 5 showing a positive relationship between circulating EV PD-L1 with residual C-peptide in individuals with type 1 diabetes.

[0031] Prior work showed variable PD-L1 expression in insulin-containing islets from recent-onset type 1 diabetes donors, correlating with increased CD8 T cell presence. In some embodiments, there may be heterogeneity in circulating EV PD-L1 in autoantibody positive donors and children with recent onset type 1 diabetes, with the highest levels in single versus multiple autoantibody positive individuals. These findings could reflect that heterogeneity in PD-L1 expression could be related to timing in the context of natural history of disease and in the presence of active autoimmunity. Elevated plasma EV PD-L1 in nondiabetic islet autoantibody positive individuals but not in those with recent-onset type 1 diabetes may suggest that changes in circulating PD-L1 may be more prominent earlier in the natural history of the disease, possibly coinciding with the importance of IFN-a during early type 1 diabetes development. A preferential expression of PD-L1 in insulin positive cells from donors with type 1 diabetes or autoantibody positivity and correlations of circulating EV PD-L1 with C-peptide may reflect that heterogeneity is linked to a protective effect on beta cell survival.

[0032] EVs are typically categorized into three size-based subpopulations. Small EVs are often termed exosomes and are formed by the fusion of the multi-vesicular endosome with the plasma membrane. Larger EVs (microvesicles and apoptotic bodies) emanate from plasma membrane blebbing. In some embodiments, PD-L1 is associated with small EVs upon IFN-aexposure. In some embodiments, PD-L1 induction may be most strongly linked (6-fold) with CD81+ EVs compared to other tetraspanins (2-3-fold in CD9+ and CD63+) in EndoC-PHl EVs. Recent work has shown that CD81 is upregulated in diabetic conditions, with expression of CD81 increased in stressed human beta cells

[0042] . Differences in islets vs. beta cells may reflect the contributions of other islet cells.

[0033] In some embodiments, there is no difference in human plasma EV or non-EV PD- L1 levels between patients or subjects with type 1 diabetes and non-diabetic controls. In some embodiments, changes in the PD-L1 / PD-1 axis may be time-dependent. For example, changes in circulating PD-L1 may be more prominent earlier in the natural history of disease. In some embodiments, the relationship between circulating EV PD-L1 and residual C-peptide in both at- risk individuals and type 1 diabetes is suggestive of a beneficial contribution of islet-derived EV PD-L1 in preserving beta cells. The absence of this correlation in non-diabetic individuals may indicate that EV PD-L1 becomes particularly relevant in the context of ongoing autoimmune attack.

[0034] EVs can also serve as natural drug delivery vehicles and therapeutic molecules, implying that PD-L1 could be encapsulated within EVs and targeted to specific cell types or tissues, potentially including beta cells or immune cells involved in the autoimmune response. A beneficial effect of PD-L1+ EVs on beta cell tolerance could theoretically be harnessed as an intervention to prevent autoimmune beta cell destruction.

[0035] The present disclosure describes a method of treating a patient or subject for type 1 diabetes comprising measuring EV PD-L1 in circulation, comparing the measurement to a baseline measurement, and if the measurement of EV PD-L1 in circulation is more than the baseline measurement, treating the patient or subject by giving them a drug for type 1 diabetes.

[0036] EXAMPLES AND METHODOLOGY

[0037] Example 1 : PD-L1 is present in extracellular vesicles associated with beta cells.

[0038] PD-L1 may be intracellularly sorted into EV cargo as observed by immunofluorescent staining of rat insulinoma cells (INS-1 cells) and EndoCpHl cells to characterize PD-L1 co-localization with CD63, CD81, and CD9, tetraspanin proteins that are EV associated surface markers. As shown in FIG. 1A and Fig ID, co-localization of PD-L1, CD63, CD81, and CD9 within cells was observed. To verify the presence of PD-L1 in beta cell EVs, small EVs were isolated from rat insulinoma cells (INS-1 cells), EndoC-PHl human beta cells, and human islets. As shown in FIGS. IB and 1C, EV isolation and purity were validatedusing transmission electron microscopy (TEM) and immunoblot analysis for characteristics EV proteins and calreticulin to ensure depletion of cell debris. As shown in FIGS. 2H-2J, EV isolation and purity were also validated using nanoparticle tracking analysis (NT A). The mean diameter of the vesicles measured by TEM and NTA showed values in the range of about 50 nm to about 200 nm (majority peaking at 94 nm), consistent with small EVs. Immunoblots confirmed the presence of PD-L1 in EVs, in association with CD9 and CD63, and the absence of calreticulin, under control conditions.

[0039] Example 2: IFN-a increases beta cell EV PD-L1 cargo.

[0040] IFN-a or IFN-y signaling increases beta cell PD-L1 levels in human beta cells. To test if IFN-a also regulates beta cell EV PD-L1 expression, INS-1 cells and EndoC-PHl were treated with 2000 U / ml IFN-a for 24 h. IFN-a led to a 1 .5-fold increase in intracellular PD-L1 expression (FIGS. 2A, 2C). To determine whether this mode of inflammatory stress increases beta cell EV PD-L1, small EVs were isolated and PD-L1 content was quantified using immunoblot. IFN- a exposure induced a two to three-fold increase in small EV PD-L1 (FIGS. 2A-2D). To determine if IFN-y or IL1-P impact beta cell EV PD-L1, EndoC-PHl cells were treated with 100 ng / ml IFN-y and / or 5 ng / ml IL-ip for 24 hours. EndoC-PHl cells and EVs demonstrated robust PD-L1 upregulation upon IFN-y treatment (~3-fold increase) but not with IL1- alone (FIGS. 2E-2F). Testing of human islets showed more than a two-fold increase in EV PD-L1 levels with IFN-a (FIGS. 2G-2I). A similar trend was observed for soluble PD-L1 levels, but overall differences were smaller and were not statistically significant (FIG. 2H).

[0041] To understand if increases in EV PD-L1 were due to increased release of EVs under IFN-a stimulated conditions, NTA was performed to determine EV quantity and size distribution. NTA demonstrated that the size distribution of the EVs was consistent with small EVs. However, no significant difference in either the particle size distribution or concentration was observed after IFN-a treatment (FIGS. 2J-2I). These findings suggest that IFN-a treatment does not increase EV size or number, but rather leads to an enrichment of PD-L1 cargo in EVs from human islets and P cell lines.

[0042] Example 3: PD-L1 protein is present on beta cell EV surface and is increased by IFN-a can bind PD-1.

[0043] PD-L1 is expressed on the surface of beta cell EVs and thus, has the potential to directly interact with PD-1 on immune cells. Surface presence of PD-L1 protein in different EV subpopulations was tested by single-particle interferometric reflectance imaging to characterizeindividual EV particles using an ExoView microarray chip coated with antibody to human tetraspanin EV markers (CD63, CD81 and CD9) from EndoC-PHl and human islets treated with or without IFN-a or IFN-y. Fluorescence imaging / interferometry quantification of surface EV PD-L1 positivity on intact EV subpopulations based on presence of membrane tetraspanins was performed. At baseline, CD63, CD81, and CD9 positive EVs showed similar percentages of PD-L1 positivity on the EV surface (-10-15%) (FIG. 4A). 24-h treatment with IFN-a or IFN- Y yielded significant increases in PD-L1 positive EndoC-PHl EVs within each subpopulation, but this was the most pronounced for CD81+ EVs (~6-fold increase with IFN-a and - 10-fold increase with IFN-y) vs. - 2-4-fold increase for CD9+ and CD63+ EVs respectively (FIG. 4C- D). In human islet EVs, IFN-a or IFN-y yielded a -1.5-fold increase in PD-L1 positive EVs within each subpopulation (FIG. 4G-H).

[0044] To test if beta cell EV PD-L1 is capable of binding to PD-1, a PD-1 / PD-L1 competitive binding kit was utilized. Decreasing concentrations of EndoC-PHl EVs were used to test the PD-1 binding and the binding was compared to a positive control (PD-L1 standard). A higher concentration of EndoC-PHl EVs blocked the reagent interaction from interfering with the signal. This interference was reversed with reduced EV concentrations in a dose-dependent manner, confirming binding with PD-L1 on the EV surface (FIG. 3C). As a negative control, the binding assay was performed using EVs isolated from HEK293 cells, which do not express PD-L1 (FIG. 3D). EVs isolated from HEK293 cells did not effectively block the reagent interaction (EC50: 0.03+0.002) (FIG. 3E). However, EVs isolated from HEK293 cells after transfection with a plasmid to express PD-L1 were able to block the interaction in a dosedependent manner with 5.4-fold higher efficacy than the EVs from wild- type HEK293 cells (average EC50: 0.005+0.001) (FIG. 3E). In aggregate, this data confirms that IFN-a increases PD-L1 loading on the P cell EV surface and that PD-L1 on P cell EVs is able to bind PD-1.

[0045] Example 4: EV PD-L1 inhibits proliferation, cytokine production and cytotoxicity’ of activated murine CD8 T cells in vitro.

[0046] Beta cell EV PD-L1 suppression of CD8 T cell activation was investigated. PD-L1 expression in NIT-1 cells, which express low levels of endogenous PD-L1, was exogenously induced to generate EVs with PD-L1 overexpression (FIG. 6A). EVs were isolated from both wildtype (WT) and PD-L1 overexpressing (OE) NIT-1 cells using sequential ultracentrifugation, and EV PD-L1 expression quantified by ELISA (FIG. 6B). To assess the impact of EV-associated PD-L1 on immune cells, splenocytes were isolated from NOD mice. Splenocytes were activated with anti-CD3 and anti-CD28 antibodies, then incubated for 72hours with NIT- 1 -derived WT or PD-L1 OE EVs, and in the presence or absence of a PD-L1 blocking antibody.

[0047] Flow cytometric analysis was performed to quantify CD8+ T cell proliferation, using CTV dye and activation, quantified with CD69, CD25, or CD44 positivity (FIGS. 5, 6C-6D). PD-L1OEEVs significantly inhibited activated CD8+ T cell proliferation (FIGS. SB, 5F, 5G) and activation (FIGS. 5C-5E, 5H-5M). Importantly, pre-treatment of the PD-L1OEEVs with an anti-PD-Ll antibody nearly abolished these inhibitory effects (FIGS. 5B-5M). PD-L1OEEVs also significantly reduced the cytotoxicity of mouse splenic cells, as demonstrated by decreased levels of IFN-y and granzyme detected in the splenocyte supernatants (FIGS. 5N-5O). Again, EV pretreatment with anti-PD-Ll antibody abrogated their ability to reduce T cell potential for cytotoxicity (FIGS. 5N-5O).

[0048] Example 5: Plasma EV PD-L1 from children with recent onset type 1 diabetes correlates with residual circulating C-peptide levels.

[0049] To test the relevance of circulating EV PD-L1 to human type 1 diabetes, plasma samples from 26 children with recent-onset type 1 diabetes and non-diabetic controls matched by age, sex, and BMI were analyzed. Demographic characteristics are displayed in Table 1. EVs and non-EV soluble proteins were eluted from 500 pl plasma using size exclusion chromatography. NTA was performed on plasma EV samples to assess differences in the overall concentration and size distribution. No difference in total particle number or size distribution existed between the two groups (p value=.7168) (FIGS. 7A-7B). The circulating EV PD-L1 and soluble PD-L1 were quantified similar soluble and EV PD-L1 levels were observed for children with recent-onset type 1 diabetes and controls (FIGS. 7C-7D). No relationships between plasma EV PD-L1 and age, or sex, or BMI percentile were detected for either group.Table 1. Demographic characteristics of study participants.Variable Non-TID controls Type 1 diabetes R value n=26 n=26Age, years 10 (6-12) 10 (5.5-12) 0.8144Male sex, % 61.54 61.54 >0.9999BMI percentile 53 (28.25-81) 37.5 (12.5-86.5) 0.3045Results are displayed as median (interquartile range). BMI- body mass index.

[0050] Next, the plasma C-peptide levels was quantified. Plasma C-peptide wassignificantly reduced in children with type 1 diabetes compared to non-diabetic controls (FIG. 7E). However, in children with type 1 diabetes, plasma EV PD-L1 positively correlated with circulating C-peptide (spearman r=0.5370; p value=.0047) (FIG. 7F, red), suggesting a link between EV PD-L1 and residual beta cells in type 1 diabetes. This relationship was not present in non-diabetic participants (FIG. 7F, blue).

[0051] The circulating levels of EV-associated and soluble PD-L1 were assessed prior to the development of type 1 diabetes. EV and soluble PD-L1 levels were analyzed from previously banked de-ide ntified plasma samples from 16 nondiabetic individuals with islet autoantibody positivity and 20 autoantibody-negative nondiabetic controls (Demographics in Table 2). Here EV PD-L1 levels were significantly elevated in individuals with islet autoantibody positivity compared to controls (FIG. 7G). Interestingly, this increase was not observed for soluble PD- L1 levels (FIG. 7H). Upon stratification based on the number of islet autoantibodies, it was observed that those with a single islet autoantibody exhibited higher levels of plasma EV PD- L1 (FIG. 7G, green squares). The relationship between circulating C-peptide levels (FIG. 71) and plasma EV PD-L1 was examined. Similar to recent-onset clinical T1D, a positive correlation was observed between plasma EV PD-L1 and circulating C-peptide (Pearson's r - 0.5061; p-value =.0455) in individuals predisposed to type-1 diabetes (FIG. 7J, green). This correlation suggests a link between EV PD-L1 and residual beta cell function in these individuals at-risk for type-1 diabetes. Notably, this relationship was not present in the control group (FIG. 7J, blue).

[0052] Example 6: Methods

[0053] Cell Culture: Rat insulinoma cells, INS-1 823 / 13 (INS-1, RRID: CVCL_7226) originally obtained from Chris Newgard (Duke University Medical Center) were grown in RPMI 1640 (Gibco, Thermo Fisher Scientific) containing 10% EV-depleted FBS with supplements. NIT-1 insulinoma cell line (CRL-2055), obtained from Erica Cai (Indiana Biosciences Research Institute, IN) were grown in DMEM (Gibco, 10313039). FBS was depleted of EVs by centrifuging for 20 h at 100,000xg. EndoC-PHl cells, obtained from Human Cell Design, Toulouse, France was cultured in ECM / fibronectin-coated plates in low glucose (5.5mM) serum free DMEM with supplements. Human islets were obtained either through the Integrated Islet Distribution Program or Alberta Islet core (n=8, electronic supplementary material [ESM]) and cultured. To model the early inflammatory milieu of type 1 diabetes, cells and human islets were exposed to 2000 U / ml human IFN- a 2a (pbl assay sci, Pestka Biomedical Laboratories, #11100) or 100 ng / mL human IFN-y (285IF100, R&D Systems) or 5 ng / mLhuman IL1-P (201LB005, R&D Systems) or cytokine mix of IFN-gamma and IL1-P and INS-1 cells were exposed to 2000 U / ml Rat IFN- a 1 (pbl assay sci, Pestka Biomedical Laboratories, #13101-1) or or 100 ng / mL Rat IFN-gamma (585IF100, R&D Systems) for 24 h. Approximately IxlO7INS-1 cells, 8xl06EndoC-PHl cells, and 500 human islet equivalents (lEQs) from the same donor or cell passage were seeded and used for each condition for all experiments unless otherwise stated.

[0054] EV isolation: For cell lines, small EVs were isolated from culture media using ultracentrifugation. Briefly, 10-15 ml of supernatant was centrifuged at 800 g for 10 minutes. The supernatant fraction was centrifuged at 2,000 g, passed through 0.22 m filter (Merck Millipore), and was centrifuged at 10,000 g for 1.5-h at 4°C to pellet small EVs. Total EV protein concentrations were determined by using a BCA protein assay kit (Thermo Fisher Scientific). Isolation and relative purity of the EVs were confirmed by nanoparticle tracking analysis (NTA), transmission electron microscopy, and immunoblot. For NTA, samples were analysed using a ZetaView instrument (ParticleMetrix) per manufacturer instructions. Particle sizes and numbers was analysed at 11 positions per sample and calculated as the mean of the results with ZetaView Analyse software (ParticleMetrix). EVs were isolated from human islet medium and 500 pl plasma using size exclusion chromatography (SEC). The samples were centrifuged at 2,000 g for 10 minutes and ultra- filtered using 1 pm filter (Cytivia, GE Healthcare distributor). After rinsing the qEV columns (Izon Science) with 0.22 pm filtered IX PBS, 500 pl of the sample was applied on top of a column and 0.5 ml fractions were collected in 1.5 ml tubes. Four EV-rich fractions (6-9) and six soluble protein-rich fractions (13-18) were pooled and analyzed for EV purity.

[0055] Nanoparticle Tracking Analysis: EV-enriched samples were analyzed for concentration and size distribution with dynamic light scattering using a ZetaView instrument (ParticleMetrix). Samples were diluted in 0.22 pm filtered PBS to a concentration within the manufacturer’s recommendations. 1 ml diluted sample was loaded into the flow cell and was recorded for 55 seconds. Particle sizes and numbers were analyzed at 11 positions per sample and calculated as the mean of the results with ZetaView Analyse software (ParticleMetrix).

[0056] Transmission Electron Microscopy (TEM): Samples were spotted onto formvar / silicon monoxide coated 200 mesh copper grids (Ted Pella Inc. Redding, CA). Grids were glow discharged for 60 seconds at 20pA with a GloQube glow discharge unit (Quorum Technologies, East Sussex, UK) before use. Samples were negatively stained with NanoVan (Nanoprobes, New York, NY) and examined on a Tecnai G2 Spirit TWIN (FEI, Hillsboro, OR)operating at an accelerating voltage of 80kV. Images were acquired digitally with an AMT (Woburn, MA) digital imaging system.

[0057] Immunoblot: Proteins were extracted from cell lysate or pelleted EV fractions using IX RIPA buffer with protease and phosphatase inhibitor cocktail (Roche). Samples were diluted using I X sample buffer (LI-COR #928-40004) with P-mercaptoethanol (Thermo Fisher Scientific). The proteins were separated using 4-20% SDS-PAGE precast gel (Biorad) and transferred to an activated PVDF membrane. Following blocking with Odyssey blocking buffer (LICOR-Biosciences #927-50003) for 45 minutes, the membrane was incubated with primary antibodies, PD-L1 (1 :1000, 17952-1-AP, Proteintech), CD63 (1:1000, ABIN144001 ; Antibodies Online), CD9 (1 :1000, 10292-1-AP; Proteintech) and calreticulin (1: 1000, 27298- 1-AP; Proteintech) overnight at 4 °C. The membrane was then incubated with IRDye 800CW and 680RD secondary antibodies (1 :10,000, LICOR Biosciences # 926-32211, # 926-68072) and imaged on the Odyssey CLX Scanner (LI-COR Biosciences).

[0058] Immunofluorescence: INS-1 cells or EndoC-PHl cells were seeded and allowed to attach overnight in a Millicell EZ slide (Millipore). Adherent cells were fixed with 4 % paraformaldehyde for 10 minutes and blocked with 2% BSA and 0.5% Triton X-100 in PBS for 45 minutes. Cells were incubated with primary antibodies against PD-L1 (1:200, 17952-1-AP, Proteintech) and CD63 (1:200, ABIN144001; Antibodies Online) or CD9 (1:200, 10292-1-AP; Proteintech) overnight at 4°C, followed by donkey anti-rabbit Texas Red (1: 1,000; Invitrogen) or donkey anti-goat 647 (1 :1 ,000; Invitrogen) and donkey anti-mouse Alexa 488 ( 1 : 1 ,000; Invitrogen) secondary antibodies. Nuclear staining was performed with a Vectasheild mounting medium with DAPI (Vector Laboratories) and a confocal microscope (LSM700, Carl Zeiss) was used for image analysis.

[0059] Flow cytometry: Briefly 1x107 of streptavidin beads (#10608D, Thermo fisher) were diluted in 1 ml PBS / 1% BSA and incubated with biotinylated CD9 or CD63 antibody at room temperature for 1 hr according to the manufacturer’s protocol. Pelleted INS-1 EVs were coincubated with antibody-coupled magnetic beads overnight at 4°C on a sample mixer. Samples were stained with APC anti-PD-Ll (clone 10F.9G2, Biolegend) and EXO-FITC (System Biosciences) according to manufacturer’s instructions and washed with 1 ml PBS / 1% BSA to remove the unbound antibodies. Samples were run on BD LSR Fortessa (X-20, BD Biosciences, CA) for 2 min and data were analyzed using FlowJo™ vl0.8 Software (BD Life Sciences).

[0060] PD-1 / PD-L1 binding assay: Homogeneous time-resolved fluorescence (HTRF) PD-1 / PD-L1 Binding Assay Kit (Cisbio Bioassays SAS, France, 64PD1PEG) was used to test the binding of EndoC-[3Hl EV PD-L1 to PD-1. HEK293 cells were transiently transfected with pGIPZ-PD-Ll-EGFP DNA (pGIPZ-PD-Ll-EGFP was a gift from Mien-Chie Hung; RRID: Addgene_120933) using lipofectamine 3000 (Life Technologies, CA, USA) for 72 hours. Following transfection, presence of PD-L1 was confirmed by Western Blot and EVs were isolated from supernatant as above. EVs from wild-type cells and HEK293 cells overexpressing PD-L1 were assayed using binding assay. Briefly, the PD1 / PD-L1 binding assays were performed in white 96-well low volume plates (Cisbio Assays) with a final volume of 20 pl comprising 2.0 pl of diluted EndoC-PHl EVs / standard, 4.0 pl of Tagl-PD-Ll (5 nM) and 4.0 pl of Tag2-PD-1 (50 nM). Following 10 min of incubation, 10 pL of pre-mixed anti-Tagl- Europium and anti-Tag2-XL665 detection reagents were added. HTRF signal was measured after 2 hours using a microplate reader (SpectraMax iD5, Molecular Devices) and the measurement conditions were set up in the SoftMax® Pro software. Results were analysed with a two-wavelength signal ratio: (intensity (665 nm) / intensity (620 nm)) xl04 (HTRF Ratio). The normalized HTRF ratio was calculated in accordance with the guidelines provided by the manufacturer: ((sample signal) - (min signal)) / ((max signal) - (min signal)) xlOO, where ‘max signal’ is the signal ratio with tagged PD-1 / PD-L1 proteins and ‘min signal’ the signal ratio without tagged PD- 1 protein.

[0061] ExoView fluorescence and interferometry imaging: PD-L1 surface staining was performed using ExoView® Exosome Human Tetraspanin Kit (EV-TETRA-C, #251-1044, Unchained Labs) with chips printed with capture antibodies for CD63, CD81 and CD9 according to the manufacturer’s protocol. Briefly, 1 x 108 -5 x 108 EVs / ml was diluted using sample incubation buffer and 50- 75 pl of samples were added to the pre-scanned chips and incubated overnight at RT. After incubation, chips were washed with solution A, and incubated with anti-CD9 (CF488, 1 :500), anti-CD81 (CF555, 1 :500), and anti-PD-Ll (FAB1562R, 1:100) for 1 hr. Following washing, the chips were imaged using ExoView R200 automated imager and analyzed using ExoView data analysis software.

[0062] Human samples: De-identified randomly collected plasma from 26 pediatric subjects with recent-onset type 1 diabetes and 26 age-, gender-, and BMI-matched non-diabetic controls were obtained from the IU biorepository at Indiana University School of Medicine. For the analysis of soluble PD-L1 34 samples (17 samples from each cohort) were assayed. Deidentified randomly collected plasma from 16 subjects with positive islet autoantibodies and from 20 autoantibody negative nondiabetic controls were also obtained from the IUbiorepository at Indiana University School of Medicine. Informed consent and assent as possible were obtained from all patients. Collections were approved by the Indiana University School of Institutional Review Board and reported investigations have been carried out following the principles of the Declaration of Helsinki as revised in 2008.

[0063] Treatment of CD8 T cells with the EVs: Murine PD-L1 , pGIPZ-mPDLl (121488, Addgene, gift from Mien-Chie Hung) were packaged into lentiviral particles using HEK 293T cells co-transfected with the viral packaging plasmids. Lentiviral supernatants were harvested 72 h after transfection. NIT-1 cells were infected with filtered lenti virus and selected by 5 pg / ml puromycin ant-pr-1, Invivogen). Mouse splenocytes of seven-to-twelve-week-old NOD / ShiLtJ mice 001976, The Jackson laboratories,) were isolated from the spleens under aseptic conditions. Individual spleens were homogenized to release splenocytes, in 5 ml MACS buffer (IxPBS, 0.5%BSA and 2mEDTA). The single-cell suspensions from individual spleens were filtered through a 40 pm cell strainer 22363547, Fisherbrand) and the erythrocytes were lysed briefly using RBC lysis buffer (R7757- 100, Sigma). The cell suspension was washed with MACS buffer and centrifuged at 500 g for 5 minutes. Cells were then stained with CellTrace Violet (CTV) (C10094, Invitrogen) as follows: a 5 mM stock solution was prepared and diluted 1:2000 in IX PBS to create a 2.5 pM working solution. After warming the diluted CTV for 5 minutes at 37°C, cells were resuspended at 1 x 106 cells / mL and incubated at 37°C in a darkened water bath for 10 minutes. The reaction was stopped by adding 5 volumes of cold medium with 10% FBS, followed by resuspension in 5 mL splenocyte medium, containing RPMI medium (Gibco) containing 10% heat-inactivated FBS (S11550H, Lot-A20005) and supplemented to a final concentration with L-glutamine (2 mM), penicillin (50 U / ml), streptomycin (50 pg / ml), 2- mercaptoethanol (50 pM), 1 % HEPES (IM, 15630-080 ), 1% Sodium pyruvate (lOOmM), 1%NEAA (11140-050) and 0.7% 2-Mercaptoethanol (M6250). Furthermore, to assess T cells mediate immune activation, 96- well round bottom plates (650180, Greiner Bio One) were coated with anti-CD3 (0.5 pg / ml, 100202, clone 17A2, Biolegend) and Purified NA / LE Hamster Anti-Mouse CD28 (1 pg / ml, 553294, BD Pharmingen™) antibodies. Splenic cells (1-1.5 x 106) were then cultured purified NITI wild-type EVs or PD-L1 overexpressing EVs with or without anti-mouse PD-L1 blocking antibody (BE0101, Bio X Cell) or Rat IgG isotype control (BE0090, Bio X Cell). CTV-labelled cells were analysed by flow cytometry at different timepoint (24 hr, 48 hr, and 72 hr) to assess CD8 T cell proliferation, and culture supernatant was collected and stored at -80 °C for cytokine analysis. Cells were analysed for activation- linked T-cell surface markers under different treatment conditions. After washing with FACSbuffer (lx PBS and 1% BSA), cells were stained using standard flow cytometry methods with eFluor 780 fixable viability dye (65-0865-18, eBiosciences), FITC anti-mouse CD45 (11-0451- 82, Thermo Fisher Scientific), PE anti-mouse CD8 (162304, BioLegend), APC anti-mouse CD25 (101910, BioLegend), BV605 anti-mouse CD44 (103047, BioLegend), and PE-Cy7 antimouse CD69 (104512, BioLegend) antibodies. Samples were recorded on Attune NxT Flow Cytometer (Thermo Fisher Scientific), and data were analysed using FlowJo™ vlO Software (BD Biosciences).

[0064] ELISA: Human islet EV, plasma EV and soluble PD-L1 levels were measured using U-plex Human PD-L1 kit (epitope 1) (Mesoscale Discovery) according to the manufacturer’s protocol. Plasma C-peptide was measured by the TOSOH immunoassay (TOSOH Biosciences). Mouse PD-L1 was detected in NIT-1 EVs using the Mouse PD-L1 DuoSet ELISA kit (R&D Systems, DY1019-05) and the DuoSet ELISA Ancillary Reagent Kit 2 (R&D Systems) according to the manufacturer’s instructions. Supernatants from the EV-Splenocyte coculture experiments were collected at indicated timepoints and assayed for Mouse Granzyme-B DuoSet ELISA (R&D Systems, DY1865-05). IFN-y ELISAs antibody pairs for mouse were purchased from BD Biosciences, performed according to manufacturer’s protocol. ELISAs were read on a SpectraMax M2 microplate reader (Molecular Devices) and data analysed using SoftMax Pro version 7.0.2 software (Molecular Devices). All samples were tested in duplicate.

[0065] Example 7: Statistics: Data were analyzed with GraphPad Prism 10.0.3 software for MacOS, GraphPad Software, Boston, Massachusetts USA, www.graphpad.com. Significance was assessed by a two-tailed Student’s t test or Mann-Whitney U test (for nonparametric distributions). Spearman’s correlation analyses were used to measure monotonic relationships. P-values were considered statistically significant when p value<0.05. Data are presented as mean ± SEM.

[0066] The figures provided herein are not necessarily to scale, although a person skilled in the art will recognize instances where the figures are to scale and / or what a typical size is when the drawings are not to scale. While in some embodiments movement of one component is described with respect to another, a person skilled in the art will recognize that other movements are possible. Additionally, a number of terms may be used throughout the disclosure interchangeably but will be understood by a person skilled in the art. Further, to the extent features, sides, or steps are described as being “first” or “second,” such numerical ordering is generally arbitrary, and thus such numbering can be interchangeable. Still further, in the present disclosure, like-numbered components of various embodiments generally have similar featureswhen those components are of a similar nature and / or serve a similar purpose. Lastly, the present disclosure includes some illustrations and descriptions that include prototypes, bench models, or experimental design. A person skilled in the art will recognize how to rely upon the present disclosure to integrate the techniques, systems, devices, and methods provided for into a product in view of the present disclosures.

[0067] While the concepts of the present disclosure are susceptible to various modifications and alternative forms, specific exemplary embodiments of the disclosure have been shown by way of example. It should be understood, however, that there is no intent to limit the concepts of the present disclosure to the particular disclosed forms; the intention is to cover all modifications, equivalents, and alternatives falling within the spirit and scope of the invention as defined by the claims. Although this disclosure refers to specific embodiments, it will be understood by those skilled in the art that various changes in form and detail may be made without departing from the subject matter set forth in the accompanying claims.

Claims

CLAIMS:

1. A method for treating type 1 diabetes in a subject comprising: measuring an amount of PD-L1 is circulation; comparing the amount of PD-L1 in circulation to a baseline value of PD-L1 ; treating the subject by providing a therapy for type 1 diabetes if the amount of PD-L1 in circulation is higher than the baseline amount of PD-L1.

2. The method of claim 1, wherein the amount of PD-L1 in circulation is comprised in extracellular vesicles formed from beta cells3. The method of claim 2, wherein the PD-L1 is located on the surface of the extracellular vesicle.

4. The method of claim 3, wherein the extracellular vesicles are exosomes.

5. The method of claim 3, wherein the extracellular vesicles are microvesicles or apop to tic bodies.

6. The method of claim 3, wherein the extracellular vesicles are CD81+ extracellular vesicles.

7. The method of claim 2, wherein number of extracellular vesicles comprising the amount of PD-L1 in circulation is not substantially different than number of extracellular vesicles comprising the baseline value of PD-L1 .

8. The method of claim 2, wherein interferon exposure increases the amount of PD-L1 in circulation to be more than the baseline value.

9. The method of claim 8, wherein the interferon is IFN-a.

10. The method of claim 8, wherein the interferon is IFN-y.1 1. The method of claim 2, wherein the amount of PD-L1 in circulation suppresses proliferation and cytotoxicity of CD8 cells.

12. The method of claim 1, wherein the PD-L1 in circulation can bind to PD-1.

13. The method of claim 1, wherein the method further comprises determining a stage of type 1 diabetes of the subject based on the amount of PD-L1 in circulation.

14. The method of claim 13, wherein a change in the baseline value is indicative of an early stage of type 1 diabetes.

15. The method of claim 1, wherein the amount of PD-L1 in circulation is measured in plasma of the subject.

16. A method of preventing autoimmune beta cell destruction comprising delivering extracellular vesicles loaded with PD-L1 to beta cells.

17. The method of claim 16, further comprising delivering extracellular vesicles loaded with PD-L1 to immune cells.

Images