Chicken‐or‐egg question: Which came first, extracellular vesicles or autoimmune diseases?

Abstract Extracellular vesicles (EVs) have attracted great interest as contributors to autoimmune disease (AD) pathogenesis, owing to their immunomodulatory potential; they may also play a role in triggering tolerance disruption, by delivering auto‐antigens. EVs are released by almost all cell types, and afford paracrine or distal cell communication, functioning as biological carriers of active molecules including lipids, proteins, and nucleic acids. Depending on stimuli from the external microenvironment or on their cargo, EVs can promote or suppress immune responses. ADs are triggered by inappropriate immune‐system activation against the self, but their precise etiology is still poorly understood. Accumulating evidence indicates that lifestyle and diet have a strong impact on their clinical onset and development. However, to date the mechanisms underlying AD pathogenesis are not fully clarified, and reliable markers, which would provide early prediction and disease progression monitoring, are lacking. In this connection, EVs have recently been indicated as a promising source of AD biomarkers. Although EV isolation is currently based on differential centrifugation or density‐gradient ultracentrifugation, the resulting co‐isolation of contaminants (i.e., protein aggregates), and the pooling of all EVs in one sample, limit this approach to abundantly‐expressed EVs. Flow cytometry is one of the most promising methods for detecting EVs as biomarkers, and may have diagnostic applications. Furthermore, very recent findings describe a new method for identifying and sorting EVs by flow cytometry from freshly collected body fluids, based on specific EV surface markers.


INTRODUCTION
This review will describe the role of extracellular vesicles (EVs) in the pathogenesis of four autoimmune diseases, namely type 1 diabetes (T1D), multiple sclerosis (MS), rheumatoid arthritis (RA), and systemic lupus erythematosus (SLE), also examining the possibility of applying them as diagnostic or therapeutic-response biomarkers, and their great potential as therapeutics.

Role and characteristics of EVs
EVs are lipid-bound vesicles released in biological fluids (e.g., blood, urine, breast milk, saliva, and cerebrospinal fluids or amniotic fluids) and in solid tissues, by almost all cell types, and increased in response to various stimuli (i.e., cell activation, apoptosis, and mechanical injury).
They represent an alternative mechanism for cell-to-cell communication and are required either to maintain tissue homeostasis and to activate a response to pathogens in the extracellular space. Although their presence within human peripheral blood had long been known, EVs were first described in 1996 as a way to exchange information between different cells, indicating their possible involvement in antigen presentation. 1,2 Studies focused on EVs have recently intensified, in order to better characterize cell-derived vesicles. EVs have been classified by size and cell-type origin; depending on their size, there are four main types of EV: (1) microvesicles (MVs) (100-1000 nm in diameter); (2) apoptotic blebs (1000-5000 nm in diameter); exosomes (20-150 nm), and multivesicular bodies (MVBs). 3 Basically, MVs and apoptotic blebs originate by outward protrusion of the plasma membrane, whereas exosomes and MVBs are generated by invagination of the plasma membrane. 4

Biological relevance of EVs
EVs are thought to function as shuttles for active molecules, such as proteins, lipids, metabolites, nucleic acids, in the exchange between different cells within an organism. 5 Depending on the nature of the body fluid and on the cell type, once EVs have interacted with target cells they can participate in immune-modulation, induce angiogenesis, promote coagulation, initiate apoptosis based on the active components carried by particles, or may control neuronal development, cellular proliferation, differentiation, and senescence (Table 1).

EVs in regulation of immune response
EVs from both immune and non-immune cells, such as mesenchymal stem cells (MSCs) and endothelial cells (ECs), contribute to antigenspecific and aspecific immune regulation. EVs exert their immunoregulatory functions through paracrine mechanisms and can promote or suppress immune responses. 6 In particular, APCs, including dendritic cells (DCs), macrophages, and B cells, regulate immune responses through direct interaction with CD4 + and CD8 + T cells and other immune cell types, such as NK and NKT cells. The activity of APCs on the immune function is mediated by several surface proteins (such as MHC class I and II costimulatory molecules and adhesion molecules), which are also present in EVs released by these cells. The current hypothesis is that the presence of immune regulatory proteins within the EVs enables APCs to modulate T cells at a distance. This regulation can be achieved through two different mechanisms: (i) a percentage of EVs remain attached to the APCs in an appropriate orientation in specific areas of their plasma membrane, so that the MHC complexes carried by the EVs are presented without further processing to cognate T cells (cross-dressing); (ii) EVs can deliver native antigens to APCs by internalization and processing, and EV-derived peptides can be used for presentation to T cells. 7

EVs in the etiology of autoimmune diseases
Genetic factors clearly predispose to the development of autoimmune diseases (AD), 8 but environmental factors are also important triggers. Epidemiologic studies strongly suggest that the rate of ADs is increasing over time, especially in industrialized countries, depending on lifestyle. In particular, nutritional patterns collectively termed the "Western diet" (WD; high-fat/cholesterol, high-protein, high-sugar, and excess salt intake) together with the frequent consumption of processed and "fast foods" have attracted interest as possible promoters of AD. Diet and the consumption of highly processed foods impact on the intestinal epithelium, which is the primary absorption interface for nutrients. In normal conditions, the intestinal epithelium functions as a barrier between host and environment, including food and microbiota. The term "microbiota" indicates the community of commensal, symbiotic, or pathogenic microorganisms that reside within the human body. In pathologic conditions, this barrier may be altered, creating a "leaky gut" that allows the passage of toxins, food antigens, as well as bacteria, into the bloodstream. In individuals with a genetic predisposition, a leaky gut may trigger the initiation and development of AD. The association between dysbiosis and AD has been characterized in-depth, but the impact of precise communities or species of microbe is not yet fully characterized, and further studies are required. However, it is interesting to note that the microbiota is shaped in the first years of life (native core microbiota) and that, once formed, it is resistant and resilient against perturbations. 9 It is noteworthy that organisms of all three kingdoms (including plants and bacteria) produce EVs. It has thus been suggested that microbial communities may secrete EVs and deliver positive or negative messages to the body (Fig. 1).  [10][11][12] Conversely, EVs may also serve as therapeutic tools for ADs: they have low immunogenicity, long half-life in circulation and, most importantly for MS and other neurodegenerative diseases, they can cross the blood-brain barrier (BBB) even without any surface modification. Thus, EVs may also be used to deliver drugs directly into the targeted organ.
The possible involvement of EVs in the treatment of autoimmune disorders is based on two different characteristics of these vesicles. First, EVs are natural carriers of functional DNA, RNA, and proteins, making them a suitable tool to deliver macromolecules and/or synthetic drugs. 13 Second, EVs can be modified on their surface markers in order to target specific tissues enhancing their therapeutic potential. 14

TYPE 1 DIABETES
T1D is a metabolic disorder characterized by elevated blood glucose levels attributed to insufficient or absent production of insulin, as a consequence of autoimmune aggression leading to -pancreatic cell loss. Long-term complications of the disease have been associated with macro-and micro-vascular problems, leading to heart diseases, stroke, blindness, and kidney disease. 15 Early diagnosis of diabetes mellitus could thus significantly improve preventive and therapeutic strategies. In this connection, there is a pressing need to identify biomarkers that could enable screening the population for the risk of developing diabetes, as well as affording a monitoring strategy for disease progression.

EVs in T1D pathogenesis
The endocrine pancreas comprises large clusters of cells known as islets of Langerhans, each containing thousands of cells. EVs contribute to the paracrine interactions among these cells, orchestrating hormonal secretion and promoting islet health and survival. 16 In the last decade, many studies have sought to characterize the nature of EVs cargo. Interestingly, in-depth analysis of their content has revealed the presence of insulin, C-peptide proteins, glutamic acid decarboxylase 65 (GAD65), low levels of glucagon and endothelial nitric oxide synthase, suggesting that they mostly originate from insulin-producing ß-cells. 17 Further, many miRNAs have been found to be enriched in EVs derived from ß-cells. 18 Regarding T1D pathogenesis, the current hypothesis is that pro-inflammatory cytokines secreted by immune cells within the pancreas contribute to creating an inflammatory microenvironment that, in turn, promotes the ß-cells attack by the immune system. 19 The exposure of these cells to inflammation promotes a primary islet inflammatory signaling, which triggers the loss of self-tolerance leading to disease onset.
Emerging evidence suggests that EVs play a crucial role in the initiation of autoimmune responses in the islets. 20 For instance, ß-cells exposed to pro-inflammatory cytokines display an altered miRNA signature compared to unstimulated cells. Further, these EVs show an apoptotic-inducing effect on recipient cells; they can also release intracellular ß-cell autoantigens (i.e. GAD65, IA-2, and proinsulin) which can be internalized by APCs resulting in autoreactive T and B cells activation. 21,22 ( Fig. 2A) Over 30 years ago, Leiter et al. discovered the presence of retroviral elements within EVs from T1D patients, and hypothesized that they could act as "mimetic antigens" to sensitize ß-cells and trigger autoimmunity. 23 Indeed, the presence of endogenous retroviruses (ERV), which are the germline-integrated remnants of ancient retrovirus infections, is detectable in pancreatic islets of T1D patients. 23 Using several monoclonal antibodies (mAbs), Tsumura et al. demonstrated that murine leukemia retrovirus (MLV) Gag and Env antigen expression is increased in non-obese diabetic (NOD) mice islets. 24 Recently, mAbs specific for MLV Env have been isolated from young NOD mice, and xenotropic ERV can be isolated from pancreatic cell lines. Taken together, these observations point to a new role for EVs in T1D as carriers of ERV particles, suggesting the possibility that autoimmunity may be triggered as a byproduct of an immune response against retrovirus-like MVs. 25

EVs as T1D biomarkers
To date, biomarkers for predicting T1D risk are susceptibility genes (especially HLA genes involved in antigen presentation) and islet autoantibodies (i.e., GAD65, IA-2, and pro-insulin). 26 However, these markers display several limitations: autoantibodies become detectable relatively late in the disease process, thus limiting their value in early disease prediction; further, IA-2 autoantibodies tend to decrease with disease duration, and insulin autoantibodies cannot be used after starting insulin therapy. 27 The identification of novel biomarkers is thus still an issue for the identification of new therapeutic targets and therapy improvement. EVs have recently been indicated as promising biomarkers. However, the active molecules present in EVs in T1D must also be understood in greater depth. It is known that the EV cargo contains miRNAs or proteins, including autoantibodies, that might be used for T1D diagnosis. Notably, islets from T1D patients have higher EV levels than healthy subjects. 25 Several EV active molecules have been indicated as T1D-inducing factors. For instance, Lakhter et al demonstrated that the serum increase of miR-21-5p in EV cargo could induce diabetes development. 28 The clinical progression of diabetes is often associated with complications such as nephropathy, retinopathy, or microangiopathy. Notably, these complications may be related to aging of the T cell system. 11 For instance, EVs originating from platelets and detected using lactadherin as vesicle marker have been found increased in T1D patients plasma with microangiopathy. 29 Increased expression of proteases in EVs (i.e., cystatin B, prostasin, and urokinase) have been found in the urine and plasma and correlated with nephropathy and retinopathy. 30 Conversely, multiple analyses of the miRNA expressed by EVs isolated from T1D patients serum/plasma indicated miR-150-5p, miR-21-3p, miR126, miR-145, and miR-30b-5p as potential biomarkers of the onset of diabetic retinopath. 31 Lastly, both an increased expression of water-channel aquaporins (AQPs), and in particular of AQP5 and AQP2, in EVs derived from epithelial tubular cells, and higher amounts of podocyte-derived EVs in the urine of T1D patients are associated with nephropathy development. 32 Despite EV's potential in the context of T1D, preclinical and clinical studies are only now gaining F I G U R E 2 EVs may trigger AD by delivering autoantigens to secondary lymphatic organs. In ADs EVs can carry autoantigens (AutoAgs) to lymph nodes and deliver them to APCs. In T1D these AutoAgs are released by pancreatic −cell (A), by oligodendrocytes in MS (B). In RA (C) and in SLE (D), EVs derived from synovial cells or apoptotic cells respectively, carry autoantigens that are recognized by autoreactive B cell, resulting in the formation of immunocomplexes importance. EV may be used as biomarker for the detection of early islet injury, and may serve to identify susceptible individuals for disease progression, before autoantibodies are detectable.

EVs for T1D therapy
Studies on NOD mice, an excellent model for studying genetic susceptibility to human T1D, show that both islet cells and MSCs exert an immunomodulatory effect by releasing highly-immunestimulatory EVs. 33 In particular, EVs derived from murine MSCs have been found to elicit a response in autoreactive T-cells and marginal zone-like B-cells. 33 and increase glomerular endothelial cell proliferation. 36 Moreover, another study showed that EVs isolated from human islets stimulate proinflammatory immune responses, and lead to peripheral blood mononuclear cell (PBMC) activation. 17 Additionally, they induce an increase in antibodies against GAD65 in PBMCs isolated from T1D patients. Furthermore, pretreatment of T1D derived PBMCs with ibrutinib, an inhibitor of Bruton tyrosine kinase that plays a crucial role in B cell maturation as well as in mast cell activation through the high-affinity IgE receptor, dampens EV-induced memory B cell activation and GAD65 antibody production. 37 Conversely, as mentioned above, EVs could be employed as therapeutic tools in the delivery of specific miRNAs. An example of this approach is the fact that the transfer of miR145 by EVs derived from bone marrow stromal cells in diabetic rats conferred neuro-restorative effects in a rat stroke model. 38 In addition to nucleic acids, EVs have been successfully used to deliver drugs such as curcumin, which is a natural polyphenol with anti-inflammatory properties, that had an effect on T1D mice after stroke, ameliorating neurovascular dysfunction. 39

MULTIPLE SCLEROSIS
MS is an inflammatory demyelinating disease of the CNS caused by autoimmune aggression against myelin proteins. 40

EVs in MS pathogenesis
The blood-brain barrier (BBB) consists of a network of endothelial cells, together with neurons and glial cells, including microglia. Disruption of the BBB is considered an important feature contributing to MS pathogenesis. 44 EVs may contribute to MS pathogenesis by spreading and amplifying CNS inflammation 45 as documented by considerable experimental evidences. However, EVs may also exert a protective and regenerative effect in the repair of injury through different mechanisms including: restoration of trophic factors, control of synaptogenesis, or removal of damaged cells. 46,47 Endothelial EVs carry metalloproteases that may promote BBB disruption, 48 molecules inducing endothelial activation, 49 53 Moreover, in MS, the endothelium/microglia crosstalk impacts on brain function. 54 Microglia-derived EVs are enriched in caspase 1, which has been shown to regulate the proteolytic activity of metalloproteases in endothelial cells. [54][55][56] Moreover, the cargo of microglia-derived-EVs can also be transferred to neurons and, through its miRNAs, silence genes involved in dendritic spine formation and synaptic stability. In vivo, injection of EVs derived from inflammatory rat microglia, enriched in miR-146a-5p, resulted in the loss of dendritic spines in neurons. 57 Furthermore, activated microglia EVs store and release interleukin (IL)-1 55 and MHC-II, propagating neuroinflammation, and providing an efficient route for rapid dissemination and epitope spreading. 58 Regulatory T cells (Tregs) play an important role in CNS autoimmune inflammation, and their function is impaired in MS. 59 Azimi et al.
showed that Tregs-derived exosomes from MS patients have a reduced capacity to suppress in vitro proliferation of effector T cells compared to those derived from healthy controls. 60 Moreover, circulating plasma exosomes from MS patients suppress the induction of Treg cells on CD4 + naïve T cells, and this mechanism is miRNA-mediated. 61 The CNS is an immune-privileged site, but since EVs can cross the BBB it may be speculated that EVs derived from brain cells might spread myelin antigens to the periphery, which would activate T cells before they enter the CNS 62

EVs as biomarkers for MS
CSF is the body fluid most closely reflecting the CNS in MS, 64 but because of ethical concerns, the majority of studies focused on searching for biomarkers in the peripheral blood, which is more readily accessible than the CSF. Nevertheless, a recent study 65

RHEUMATOID ARTHRITIS
RA is a chronic autoimmune disease, characterized by synovitis and joint damage, leading to loss of function and increased morbidity and mortality. 81,82 It is the result of a complex dysregulation of the immune system that involves both innate and adaptive immunity, characterized by chronic inflammation and development of autoimmunity. 83 In recent decades, improved knowledge of the mechanisms underlying the pathogenesis of RA has enabled potential targets for disease treatment to be identified, leading to the development of novel target therapies. These have had a groundbreaking effect on RA's natural history. However, the pathogenesis is still poorly known, and predictive models enabling treatment strategy to be tailored on an individual basis are lacking.

EVs in RA pathogenesis
In recent years, a variety of evidence has pointed towards a potential pathogenetic role of EVs in RA. Firstly, EVs are a potential source of autoantigens (Fig. 2C), which is particularly relevant in the specific setting of RA, since this disease is characterized by the development of typical autoantibodies, including anti-citrullinated protein (ACPA) and anti-rheumatoid factor antibodies. 84

EVs and RA biomarkers
EVs have been postulated as potential biomarkers in RA: both the total plasma concentration of EVs and the concentration of specific subsets of EVs have been used in the past. In the latter case, some studies have analyzed the potential role of specific subsets deriving from a defined have all been proposed as potential predictors of response, 114 including in patients treated with anti-TNF/cDMARD. However, these data are very recent and validation on larger cohorts will be required.
Interestingly, a specific plasma miRNA signature (miR-23 and miR-223) has been identified that may serve as predictor and biomarker of response to anti-TNF /DMARDs combination therapy. 115 Specific miRNAs have also been identified in RA patients as predictors of response to adalimumab or to etanercept, which are both monoclonal antibodies that inhibit TNF-. 116 Conversely, Krintel et al.
identified the combination of low whole blood expression of miR-22 and high expression of miR-886.3p as a predictor of good response to EULAR. 117 Lastly, in a further study, an elevated level of plasma miR-27a-3p before treatment was significantly associated with remission at 12 months, in a group of patients treated with adalimumab and methotrexate. 118 It is evident that these findings are contradictory and that, although promising, the use of the EV signature for treatment allocation is still far from being clinically applicable. However, the preliminary results obtained on discovery cohorts deserve further validation in the near future, to better understand the potential usefulness of EV characterization in clinical practice.

EVs for RA therapy
In the context of autoimmune disorders, RA is a perfect model for Lastly, EVs can be used as liposomes, to deliver a specific drug directly to the joint. In 2014, a synovium-specific targeted liposomal drug loaded with glucocorticoids was employed to specifically target FLS and endothelial cells. This strategy was effective in vivo, suppressing the inflammatory response in affected joints, 127 and confirming that EVs are a promising vehicle for drug delivery in RA. Clinical prognosis and treatments have improved over time, and treatment chiefly comprises the use of steroidal and nonsteroidal anti-inflammatory drugs, immunosuppressive agents, and biologic agents. 129 However, clinicians still lack biomarkers for prediction of disease outcome, and several studies are aimed at addressing this issue.

EVs in SLE pathogenesis
Anti-ANAs and anti-DNA antibodies are associated with the severity of SLE. 130 ANAs can bind to DNA, RNA, proteins in the nucleus, and form immune complexes (ICs) of nucleic acids. ICs constitute the serological hallmark of SLE and can drive its pathogenesis by depositing within target tissues, in particular the kidney. 130 ICs settle in the tissues, where they either to fix complement or induce cytokine production (most prominently type 1 interferon), thus inciting inflammation.
Nucleic acids usually associate to proteins both inside and outside the cell. 131 Thus, the loss of B cell tolerance to DNA and/or chromatin represents a major mechanism of SLE pathogenesis. Moreover, anti-DNA antibodies can also cross-react with other self-antigens 132 (Fig. 2D). EVs

EVs and SLE biomarkers
Autoimmunity is characterized by cell activation, and cell stimulation leads to the shedding of phosphatidylserine (PS)-rich EVs, which may be suitable biomarkers for SLE and other ADs. PS may thus be used to detect EVs in the body fluids. 139 In SLE, autoantibodies to chromatin, including nucleosomes, usually serve as especially sensitive biomarkers of the disease. 140 Recently, using the proteomics approach, Ostergaard and colleagues showed that the cargo of plasma circulating EVs in SLE patients contained an increase of specific proteins (e.g., complement, IgG, microtubule proteins, fibronectin, and desmosomal) compared to healthy controls, and these expression patterns correlate with the disease progression. 141 Circulating EVs were detected in the plasma of SLE patients using multiparametric flow cytometry. These studies have outlined novel subpopulations of platelet, endothelial, and leukocyte-derived EVs, some of which have clinical and serological correlations. In particular, frequencies of platelet-derived EVs (characterized as CD41 + , and CD41 + -CD40L + ), were found to be increased compared with healthy controls, regardless of disease activity. 108 Furthermore, endothelial activation and damage is commonly observed in SLE patients, and is related to the development of nephropathies and vascular diseases. In this connection, it has been found that endothelial-EVs (detected as CD144 + , VCAM + ) and tissue leukocyte-EVs (TF + -CD45 + ) are highly up-regulated compared to healthy controls, and these markers directly correlate with the degree of inflammation, but also glomerulonephritis and vascular dysfunction. 141 In addition, they have been proposed as markers of cytoskeletal composition defects. 141 The protein signature of endothelial-EVs could also be used as biomarker of the activity and progression of SLE, and of the presence of possible complications. As mentioned above, one of the most affected organs in SLE is the kidney: the possibility to assess glomeru- Furthermore, in mice with acute kidney injury, treatment with MSC-EVs resulted in the prevention of chronic tubular inflammation and of renal damage, and also prolonged survival. 149 Intriguingly, the therapeutic potential of MSC-EVs has been tested in a human clinical trial: a cohort of patients with chronic kidney disease were recruited and subjected to the administration of autologous MSC-EVs. Analyzing urinary parameters (i.e. blood urea, serum creatinine, urinary albumin creatinine ratio, and estimated glomerular filtration rate) they monitored kidney functions. Lastly, to evaluate the treatment-induced improvement of inflammatory immune activity, they evaluated the following parameters: TNF-, TGF-1, and IL-10, which are all involved in immune-regulation and inflammation.
Interestingly they discovered that treated MSC-EVs patients exhibited significant increases in plasma levels of TGF-1, whereas TNF-was significantly decreased, suggesting amelioration of the inflammatory immune reaction. 150

TECHNICAL ISSUES
EVs have attracted great interest as important contributors to the autoimmune response, not only at its onset, but also for their immunomodulatory activity, and their potential as biomarkers for disease activity or response to therapy. However, data from published studies are very variable, being biased by the difficulties involved in EV isolation and characterization.
It is still challenging to isolate circulating EVs with good recovery and without contamination from proteins and lipoproteins. To date, most isolation protocols are based on differential centrifugation. However, high velocities generate protein and vesicle aggregates. Further, isolation of vesicles from plasma or serum by density-gradient ultracentrifugation results in co-isolation of high-density lipoprotein (HDL), and isolation of HDL results in co-isolation of vesicles. There is thus an urgent need for a simple and fast protocol to isolate vesicles from human samples. Determining the optimal strategy for isolating EVs is a critical step toward retrieving the maximal amount, while ensuring the recovery of all different vesicular subtypes and subpopulations, including the rare ones. 151 According to the recent MISEV2018 guidelines, 152 several methods have been proposed for the detection, quantification, and characterization of EVs. These include differential ultracentrifugation, sizeexclusion chromatography, immunoaffinity capture, microfluidics, and the use of exosome commercial kits. However, different methods of isolation might lead to different outcomes or discordant results, even starting from the same source. Another important issue to be addressed is that, during manipulation of biological samples (i.e. from blood to serum), platelets and other cells may release EVs owing to their damage, activation, or disruption. As a consequence, direct evaluation on the blood would be preferable. In this connection, flow cytometry is one of the most promising methods. Indeed, this technology can be used to characterize EVs in liquids, including blood, plasma, and other biologic fluids, as well as in solid tissues. One of the main advantages of cytometry is the possibility to detect rare populations and distinguish them from abundant ones. Moreover, the technology enables the populations to be individually sorted, and they can then be typed in depth through 'omic' analyses (proteomics, transcriptomics, metabolomics). Interestingly, very recent studies described a new method for the identification and sorting of EVs by flow cytometry, using a lipophilic cationic dye that diffuses through plasma membranes and directly binds to EVs (Pan-EV dye) in unmanipulated body fluids (tears and CFS) in combination with specific EVs surface markers. 65,153,154 Further experiments will be needed to validate this method in the AD setting, and compare the results with those obtained using conventional procedures accepted by the scientific community.

CONCLUDING REMARKS
Currently, there are still significant gaps in our knowledge of the mechanisms underlying the pathogenesis of autoimmune diseases, in particular during the prodromal phases, and few initiating autoantigens have been identified. Indeed, ADs are usually only diagnosed sometime after epitope spreading, a process during which the immune response, begun by the autoantigen, is taken over by new T or B cell specificities. This makes it difficult to identify the "initiating antigen".
Such identification would be of great utility, since preclinical studies have revealed a window of therapeutic opportunity before the overflow of epitope spreading. Identifying the starting antigen would help to fully understand the etiology and the pathogenesis of autoimmune diseases, and to develop better immunotherapeutic approaches. In the opinion of the reviewers, a great revolution in the search for AD initiating autoantigens will come from studying rare subpopulations of EVs. Alongside the mechanisms that have been described for individual diseases (T1D, MS, RA), an initiating role may be ascribed to molecular mimicry, in which a foreign antigen (usually a viral antigen) shares sequence or structural similarities with self-antigens. This concept