Immune cell-derived membrane nanovesicles: A promethean fire for autoimmune disease therapy through immune cell mimicry Authors: Yan Li, Nasrullah Jan, Jimin Zhang, Yang Qin, Gang Liu, Chao Liu Published: 2026 ABSTRACT AbstractAutoimmune diseases (AIDs) constitute a heterogeneous group of disorders characterized by immune dysregulation, loss of self-tolerance, and chronic inflammation, which leads to tissue damage and organ dysfunction. Current therapies for AIDs are often limited by their lack of specificity, systemic side effects, and insufficient restoration of immune tolerance. Recent advances in nanotechnology and bioengineering have introduced immune and associated cell-derived membrane vesicles (IACMVs) as a promising therapeutic platform. Derived from macrophages, dendritic cells, neutrophils, platelets, or red blood cells, IACMVs inherit key surface proteins and receptors from their parent cells, conferring endogenous biocompatibility, inflammation-specific targeting, and intrinsic immunomodulatory capabilities. These vesicles can be engineered to carry therapeutic cargoes (e.g., peptide inhibitors, nucleic acids) or modified with surface ligands to enhance disease-site specificity, making them versatile tools for specific immunomodulation. This review provides a comprehensive overview of IACMVs, focusing on their preparation techniques, functional mechanisms, and therapeutic applications in prototypical AIDs such as rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), inflammatory bowel disease (IBD), autoimmune hemolytic anemia (AIHA), type 1 diabetes (T1D), multiple sclerosis (MS), and autoimmune myocarditis (AM). We highlight translational challenges, including production scalability, membrane integrity, immunogenicity, and cargo-loading efficiency, that must be addressed to advance clinical translation. Finally, we discuss future directions for optimizing IACMVs as next-generation, safe, and targeted immunotherapeutic platforms for AIDs.Keywords: Autoimmune diseases, Immune and associated cell-derived membrane vesicles, Nanomedicine, Targeted therapy, Biomimetic nanoparticles Introduction Autoimmune diseases (AIDs) are prevalent chronic conditions encountered in clinical practice [1], affecting approximately 5-10% of the global population. These diseases are more prevalent in women than in men [2,3]. Current treatment for AIDs faces significant limitations. Existing therapies are often broad-acting and non-specific, resulting in various side effects [4,5]. Substantial progress has been made in the clinical management of AIDs in recent years. The emergence of biologic drugs and small-molecule inhibitors has greatly advanced the treatment of multiple typical AIDs, including rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), inflammatory bowel disease (IBD), autoimmune hemolytic anemia (AIHA), type 1 diabetes (T1D), multiple sclerosis (MS), and autoimmune myocarditis (AM) [6,7]. Nevertheless, current therapeutic strategies still face prominent limitations and unmet clinical needs regarding therapeutic efficacy and safety. Conventional corticosteroids and immunosuppressants, as the mainstream clinical regimens, can relieve inflammatory symptoms to a certain extent, but they often induce severe systemic side effects, including increased infection risk, endocrine and metabolic disorders, osteoporosis, gastrointestinal toxicity, and even elevated malignancy susceptibility [4,[8], [9], [10]]. Moreover, most targeted interventions fail to restore long-term immune tolerance and cannot achieve sustained drug-free remission, while immunotherapies such as B-cell depletion also exhibit incomplete response rates and individual patient variability [1,[11], [12], [13], [14]]. Therefore, there is an urgent demand for safer, more precise, and tolerable therapeutic strategies to address unmet clinical needs in autoimmune disease treatment. Recent advances in nanotechnology and bioengineering have promoted the rapid development of cell membrane-based biomimetic platforms, which inherit native cellular biological functions and overcome the limitations of synthetic nanomaterials [15,16]. Early cell membrane biomimetic studies mainly focused on tumor-derived membrane vaccines for antitumor immunotherapy [17,18]. By virtue of their excellent biocompatibility and endogenous bioactivity, cell membrane-mimetic strategies have been further extended to autoimmune disease therapy, giving rise to IACMVs. Distinct from tumor vaccines that initiate immune activation, IACMVs inherit native surface receptors, adhesion molecules, and immune regulatory components from parental cells, enabling targeted accumulation in inflamed tissues and effective modulation of aberrant immune responses. Given the widespread conceptual ambiguity surrounding biomimetic nanocarriers, we established a clear classification system for five typical vesicle types in Table 1 to unify the definition and research scope of IACMVs. These categories include natural extracellular vesicles (EVs), plasma-membrane-derived nanovesicles, single and hybrid membrane-coated nanoparticles, as well as synthetic vesicles with biomimetic surface engineering. While earlier reviews mostly overview these carriers as a collective group without detailed differentiation, we systematically compare their structural origins and functional features. This well-defined framework not only standardizes fundamental concepts, but also differentiates our review from previous publications. Table 1Classification and characteristics of different biomimetic vesicle and nanoparticle platforms.Table 1CategoryCore DefinitionStructural OriginMain FeaturesReferencesNatural EVsEndogenously secreted lipid bilayer vesicles released by living cells under physiological or pathological conditionsSecreted naturally from parental cellsPossess complete native biological functions; heterogeneous in size and composition; limited manual modification[[19], [20], [21]]Plasma-membrane-derived nanovesiclesNanoscale vesicles directly extruded or prepared from isolated cell plasma membranesPurified cell plasma membraneRetain surface proteins and lipid components of source cells; higher yield than natural EVs[[22], [23], [24]]Cell membrane-coated nanoparticlesInorganic/organic core nanoparticles wrapped with a single type of complete cell membraneSynthetic nanoparticle core + single cell membraneCombine the physicochemical properties of nanoparticles and biological functions of cell membranes[[25], [26], [27]]Hybrid membrane-coated nanoparticlesNanoparticle cores coated with mixed membranes from two or more distinct cell typesSynthetic nanoparticle core + multiple cell membranesIntegrate multiple biological tropisms and functions from different cell sources[[28], [29], [30]]Synthetic vesicles with biomimetic surface engineeringArtificial lipid vesicles modified with biomimetic proteins, ligands or functional peptides on the surfaceFully synthetic lipid materialsHigh uniformity and reproducibility; artificially constructed biomimetic functions[20,31] Classification and characteristics of different biomimetic vesicle and nanoparticle platforms. On this basis, we further refine the definition of IACMVs by distinguishing them from naturally secreted exosomes, two morphologically similar vesicle subtypes that are frequently confused. Specifically, IACMVs are fabricated via mechanical disruption or homogenization of isolated parental plasma membranes [32], whereas natural exosomes are endogenously derived from intracellular endosomal multivesicular bodies and released through canonical exocytosis [33]. The two vesicles differ fundamentally across multiple dimensions. Artificially prepared IACMVs fully reserve source-cell membrane components, providing inherent inflammatory targeting, controllable immunogenicity, and convenient post-modification for scalable production. In contrast, naturally secreted exosomes contain enriched endosomal proteins but incomplete plasma membrane constituents, resulting in poor intrinsic targeting, variable immunogenicity, limited engineering feasibility, and batch fluctuation restricts their clinical translation. These critical distinctions consolidate the precise research boundary of IACMVs [[34], [35], [36]]. Retaining complete native membrane components, IACMVs exert dual targeting effects: they actively home to inflammatory lesions via intrinsic surface functional molecules and also accumulate at pathological sites passively under local microenvironmental cues [37,38]. Their active homing performance can be further strengthened through membrane engineering strategies, such as targeted receptor overexpression. For example, CXCR4-overexpressing endothelial membrane-coated nanoparticles loaded with stem cell secretome that achieve efficient active targeting to bone lesions. This represents a significant advancement, shiftingthe field from passive accumulation toward highly specific active targeting of biomimetic nanocarriers [39]. With intact membrane structures, IACMVs can interact with pathological immune microenvironments, scavenge pro-inflammatory mediators and deliver therapeutic payloads with high precision [40], and their immunomodulatory functions are realized through three synergistic pathways: adsorbing inflammatory factors via decoy receptors, competitively blocking aberrant immune signaling, and modulating inflammatory cascades via intracellular delivery of therapeutic cargos [[41], [42], [43]]. Furthermore, IACMVs serve as highly adaptable modular platforms that can be functionally engineered to optimize therapeutic performance against diverse autoimmune disorders. Functional modifications, including therapeutic cargo loading (peptide inhibitors, nucleic acids), surface ligand conjugation, protein engineering, glycan remodeling, and stealth modification, significantly enhance their targeting accuracy, immune stability, and therapeutic efficacy [[44], [45], [46]]. The integration of stimuli-responsive cores also enables intelligent on-demand drug release in inflammatory microenvironments [47]. IACMVs are primarily derived from the plasma membranes of typical immune cells, including macrophages, dendritic cells (DCs), and neutrophils [[48], [49], [50], [51]]. In this review, IACMVs refer to membrane vesicles featuring biological properties related to autoimmune inflammation, immune regulation, and in vivo circulation. In addition to canonical immune cell-derived vesicles, RBC-and platelet-derived vesicles are included owing to their membrane-derived immune evasion and long-circulation abilities, which enhance vesicles in vivo persistence and inflammatory lesion accumulation. Although tumor cell and mesenchymal stem cell (MSC) membranes are not the main research focus herein, their unique immunomodulatory and biomimetic properties provide valuable design references for the future engineering optimization of IACMVs. Given the above versatile biological characteristics and engineering potential of IACMVs, this review systematically summarizes the latest progress of IACMVs in autoimmune therapy, focusing on their fabrication strategies, immunomodulatory mechanisms, engineered functional optimization, and current applications in typical AIDs. We also discuss the key translational challenges, including production scalability, membrane integrity, immunogenicity, and cargo-loading efficiency. This review aims to provide comprehensive insights and technical references for the development of next-generation safe and precise biomimetic immunotherapeutic platforms for autoimmune diseases. Pathophysiology and immunological targets in autoimmune diseases Fig. 1 demonstrates the mechanisms of immune dysregulation and therapeutic targeting in AIDs. The following sections will provide a brief discussion of these mechanisms and their therapeutic implications. Mechanisms of Immune Dysregulation and Therapeutic Targeting in Autoimmune Diseases (A) Adaptive Immune Dysregulation: Dendritic cells activate T cells, which drive B cell-mediated autoantibody production and immune complex-induced tissue damage. (B) Innate Immune Activation and Chronic Inflammation: Neutrophils and macrophages fuel chronic inflammation (e.g., in RA, via NF-κB translocation and cytokine release). (C) Defective Apoptotic Cell Clearance: Impaired macrophage phagocytosis leads to secondary necrosis, autoantigen release, and disrupted immune tolerance. (D) Therapeutic Targeting of Immune-Microenvironment Crosstalk: Strategies include modulating T cell subsets, cytokine signaling, endothelial activation, and leveraging immune-derived vesicles/chemokine axes to promote anti-inflammatory repair. All schematics in this figure are original and self-drawn. Breakdown of immune tolerance and adaptive immune dysregulation AIDs fundamentally arise from a breakdown in immune tolerance, allowing autoreactive T and B lymphocytes to persist and propagate inflammation [45,52]. DCs, as key antigen-presenting cells, aberrantly present self-antigens via major histocompatibility complex molecules, initiating T cell priming [49]. In SLE, a critical pathway involves CD40-CD40L interactions between T follicular helper cells and B cells, which promote germinal center formation and stimulate the production of high-affinity autoantibodies such as anti-double-stranded DNA (dsDNA) and anti-Smith antibodies [45,53]. These autoantibodies form immune complexes that deposit in tissues and activate complement cascades, particularly affecting the kidneys in lupus nephritis [45,54]. Moreover, autoantibodies are not limited to passive roles. In type II hypersensitivity diseases-such as AIHA, immune thrombocytopenia (ITP), myasthenia gravis, and Graves’ disease-they directly bind to surface or matrix antigens, inducing complement-mediated lysis or disrupting receptor signaling [55,56]. The presence of autoreactive antibodies that function as receptor agonists or antagonists can also interfere with critical signaling pathways in target cells [46,55]. These processes illustrate how adaptive immune dysregulation, driven by persistent antigen stimulation and defective regulatory checkpoints, becomes self-sustaining and destructive [45,52]. From a therapeutic perspective, the aberrant adaptive immune activation dominated by DC antigen presentation, T-B cell crosstalk, and autoantibody production provides promising intervention targets for IACMV-based strategies. In particular, DC-derived and macrophage-derived MVs can be engineered to modulate antigen presentation, block pathological T-B cell interactions, and restore self-tolerance, thereby reversing adaptive immune dysregulation in AIDs. Innate immune activation and chronic inflammation Innate immunity plays a crucial role in amplifying and sustaining autoimmune inflammation [46,47,57]. Macrophages and neutrophils are key effectors in this process. In IBD, RA, and SLE, macrophages exhibit a skewed polarization toward the pro-inflammatory M1 phenotype, producing tumor necrosis factor-alpha (TNF-α), interleukin-6 (IL-6), IL-1β, and reactive oxygen species (ROS), thereby exacerbating epithelial or joint tissue injury [58,59]. Conversely, M2-like macrophages, which are associated with resolution and tissue repair, are notably underrepresented. Modulating this polarization shift is increasingly being explored as a strategy to restore immune dynamics in SLE and joint autoimmunity [47,59]. Neutrophils contribute to tissue damage through infiltration, degranulation, and the release of neutrophil extracellular traps (NETs), which expose autoantigens and perpetuate inflammation [57,60]. Their interaction with the vascular endothelium via intercellular adhesion molecule-1 (ICAM-1) and vascular cell adhesion molecule-1 (VCAM-1) facilitates transmigration and amplify cytokine production [52,60]. In RA and vasculitis, neutrophil hyperactivity correlates with disease flares and tissue destruction [57,60]. At the molecular level, innate immune sensors such as TLR7 and TLR9 activate the MyD88 signaling cascade, resulting in the nuclear translocation of NF-κB and the expression of pro-inflammatory genes [44,46]. Park et al. developed a MyD88-inhibitory platform that disrupts these pathways, demonstrating promise in preclinical models of RA and SLE [44]. Cytokine-targeting biologics, including TNF inhibitors and IL-6 receptor blockers, have shown substantial clinical benefit; however, variable patient responses and the risk of immunosuppression remain concerns [46,47]. Given the central roles of macrophage polarization imbalance, neutrophil-mediated tissue injury, and persistent TLR/MyD88-driven innate inflammation, IACMVs derived from neutrophils and macrophages offer unique advantages for targeted innate immune modulation. These cell-mimetic vesicles can selectively accumulate at inflamed lesions, normalize macrophage polarization, scavenge pro-inflammatory mediators, and suppress persistent innate immune cascades to alleviate chronic autoimmune inflammation. Impaired resolution mechanisms and apoptotic clearance defects Effective immune resolution requires the timely and efficient clearance of apoptotic cells-a process disrupted in several autoimmune conditions, particularly SLE [53,54]. When apoptotic cell debris is not removed, it undergoes secondary necrosis, releasing nuclear and membrane components that act as autoantigens [61,62]. Kato et al. reported that apoptosis-derived membrane vesicles (AdMVs), enriched with dsDNA, circulate in the serum of SLE patients and activate the cGAS-STING-IFN type I pathway [53]. This activation perpetuates immune cell stimulation and exacerbates systemic inflammation. Moreover, plasma membrane vesicles (PMVs) released during apoptosis interfere with efferocytosis by competing for binding to phosphatidylserine receptors on macrophages [54]. This competition results in inefficient clearance and increased exposure of autoantigens, particularly phospholipids and chromatin, thereby promoting the formation of anti-phospholipid antibodies and further immune complex formation [45,54]. These findings suggest that impaired apoptotic clearance is not merely a downstream effect but a central pathogenic mechanism in SLE and potentially other systemic autoimmune disorders. Therapeutic strategies, including DNase therapy, type I IFN blockade, and engineered vesicle-based immune modulators, are currently under investigation to restore this defective pathway [49,53,54]. Since defective apoptotic clearance and excessive self-antigen exposure are core pathogenic drivers of systemic autoimmunity, biomimetic IACMVs can be rationally designed to restore immune resolution. Specifically, stealth-modified or macrophage-mimetic vesicles can enhance efferocytosis, scavenge circulating pathogenic apoptotic vesicles, and inhibit cGAS-STING-mediated interferon responses, thereby blocking autoantigen-driven inflammatory loops. Systemic immune-microenvironment crosstalk and therapeutic implications Autoimmune inflammation extends beyond the primary tissue targets, reflecting systemic dysregulation of immune-stromal communication. Monocytes and macrophages under inflammatory stress release chemokines such as MCP-1 and CCL2, which promote further immune cell recruitment and endothelial activation [47,52,63]. In response to cytokine exposure, the endothelium upregulates adhesion molecules (e.g., ICAM-1, VCAM-1), facilitating leukocyte extravasation into target organs [64,65]. This crosstalk exacerbates not only tissue damage but also vascular pathology, accounting for high incidence of comorbid atherosclerosis in patients with RA and SLE [66,67]. Stromal cells, including epithelial cells and fibroblasts, play a pivotal role in the chronicity of autoimmune inflammation. By secreting matrix-degrading enzymes, cytokines, and pro-survival signals, these cells promote the retention and effector functions of lymphocytes, thereby exacerbating and sustaining the inflammatory response [[68], [69], [70]]. Under chronic inflammatory conditions, stromal cells undergo phenotypic and functional changes that confer pathogenic properties. These altered stromal cells establish positive feedback loops with immune cells such as T and B lymphocytes, further driving disease progression [68,71]. For example, Th17 cells can activate synovial fibroblasts via IL-17 signaling, inducing the production of GM-CSF, which contributes to both the initiation and persistence of arthritis [69,72]. Additionally, stromal cells secrete chemokines such as CCL2 to recruit myeloid-derived suppressor cells, thereby modulating the local immune response. This modulation can lead to either enhanced inflammation or, potentially, immunosuppressive and therapeutic effects [73,74]. In AIDs such as RA and SLE, the complex interactions between stromal and immune cells are considered major drivers of disease chronicity and therapeutic resistance [75,76]. Consequently, targeting stromal cells and their crosstalk with immune cells is emerging as a promising therapeutic strategy in AIDs [[68], [69], [70]]. Given the spatial and cellular complexity of autoimmune pathology, biomimetic strategies that leverage the natural targeting properties of IACMVs are particularly promising [52,57]. Vesicles derived from neutrophils, macrophages, or DCs retain key surface markers and chemokine receptors, enabling them to home to inflamed tissues and interact selectively with immune and stromal components [49,59]. Furthermore, these vesicles can be engineered to deliver anti-inflammatory agents, immune checkpoint ligands, or RNA therapeutics, offering site-specific, low-immunogenicity modulation of immune responses. As Qiu et al. suggest, such vesicles represent a biologically intelligent system for modulating disease progression at the immune-tissue interface [57]. Types of cell membrane vesicles and their properties Macrophage-derived vesicles Macrophage-derived membrane vesicles (MMVs) are among the most extensively studied IACMVs due to their crucial roles in innate immunity, tissue homeostasis, and the regulation of inflammation [77]. These vesicles retain key surface proteins from their parent cells, including integrins (e.g., α4β1), CD14, scavenger receptors, and CD47, which facilitate targeting to inflamed endothelium, evasion of immune clearance, and sequestration of pro-inflammatory molecules within diseased microenvironments [77]. Notably, MMVs can be subclassified based on the polarization state of the source macrophages: vesicles derived from M1 macrophages exhibit pro-inflammatory and antimicrobial functions, whereas M2-type vesicles are enriched with immunoregulatory ligands that promote tissue repair and the resolution of inflammation [59,77]. Building on this classification, M2 macrophage membrane-coated nanocarriers retain the intrinsic anti-inflammatory phenotype of their parent cells and exhibit inherent tropism for inflamed tissues [59]. These vesicles can effectively scavenge pro-inflammatory cytokines, providing localized immunomodulation without causing systemic immunosuppression [59]. Macrophage-coated nanoparticles engineered to display cytokine receptors, such as TNFR and IL-6R, can act as natural cytokine decoys, binding excess inflammatory mediators and attenuating immune overactivation [46]. This functional versatility, rooted in preserved membrane composition, enables MMVs to interact with and modulate complex inflammatory networks [47,77]. Macrophage-derived vesicles can be engineered for diverse therapeutic loading and delivery purposes, leveraging their natural biocompatibility and inflammation-targeting capacity [78]. The preserved membrane profile allows these vesicles to maintain stability in biological environments and undergo site-specific accumulation within inflamed microenvironments [78]. Their functional adaptability enables integration with responsive release systems, further enhancing localized delivery and therapeutic precision [79]. In a broader context, macrophage-mimetic nanoparticles (MM-NPs) fabricated from activated macrophage membranes carry receptors such as CD11b, CCR2, and TLR4 [47,77]. These receptors enable the nanoparticles to home to inflamed vasculature and recognize inflammatory and pathogenic signals. This receptor repertoire supports applications in both infection-related and sterile inflammatory settings, highlighting MMVs as a versatile platform for modulating diverse immunopathological conditions, including autoimmune disorders. DC-derived vesicles DC-derived membrane vesicles (DCMVs) inherit the hallmark immunological features of their parental cells-most notably, their capacity for antigen presentation. These vesicles retain MHC class I and II molecules, co-stimulatory molecules such as CD80 and CD86, and intercellular adhesion molecules, enabling them to modulate immune responses toward either activation or tolerance, depending on their engineered content and context [49]. This dual potential makes DCMVs highly valuable in regulating immune homeostasis, where either immunosuppression or immune priming may be required [49,80]. Recent advances in DCMV generation have significantly enhanced their translational potential. Nitrogen cavitation facilitates the production of DCMVs with preserved membrane protein integrity and a uniform size distribution, thereby improving both pharmacokinetic properties and manufacturing consistency [49]. Moreover, targeted delivery can be achieved through aptamer conjugation, enhancing specificity while minimizing off-target immunogenicity [80]. In addition to presenting antigens, DCMVs can be loaded with immunosuppressive agents or RNA therapeutics, combining immune specificity with therapeutic payloads. Notably, their capacity to promote regulatory T cell (Treg) induction addresses fundamental immune imbalance in chronic inflammation [80]. A significant advancement in this field is the development of epigenetically reprogrammed dendritic cell-derived nanovesicles (mDCNVreg) by Xi et al. [81]. These vesicles retain essential immunological functions, including antigen presentation through elevated MHC class II expression, while simultaneously suppressing co-stimulatory signals (CD80/CD86) and enhancing inhibitory signals (PD-L1). Importantly, their tolerogenic phenotype is stabilized via IFN-γ mediated epigenetic programming, avoiding phenotypic instability observed in live tolerogenic DC therapies. Functionally, mDCNVreg efficiently traffics to lymphoid organs and induces antigen-specific CD4+ T cell anergy, while also simultaneously promoting Treg differentiation [81]. Neutrophil-derived vesicles Neutrophil-derived membrane vesicles (NMVs), an important subset of IACMVs, are biomimetic vesicles fabricated from intact neutrophil plasma membranes. Owing to the retention of native membrane components, NMVs naturally home to sites of acute inflammation. Notably, abundant surface biomarkers such as LFA-1 (CD11a/CD18) integrins specifically bind to endothelial ICAM-1, serving as the key molecular determinant for targeted accumulation at inflammatory sites [52,60]. Apart from adhesion integrins, preserved proteolytic enzymes, chemokine receptors, and ROS-related enzymes on the vesicle surface also function as characteristic biomarkers. These biomarkers enable NMVs to bidirectionally tune inflammatory cascades toward either inflammatory aggravation or resolution depending on local pathological microenvironment [52,60]. From the physicochemical perspective, standardly prepared NMVs typically possess a suitable nanoscaled size range, narrow particle size distribution and mildly negative surface zeta-potential [60]. These inherent physicochemical properties are shared among most neutrophil-derived biomimetic vesicles. Auniform size distribution effectively prevents heterogeneous in vivo behaviors triggered by uneven membrane composition, while moderate negative surface charge minimizes undesired electrostatic interactions with circulating plasma proteins. Collectively, these intrinsic features greatly decrease random phagocytic uptake by reticuloendothelial cells, reduce unexpected complement cascade activation and off-target inflammatory reactions, and thereby endow NMVs with favorable biosafety and systemic immune compatibility for in vivo administration [82,83]. Functionally, retained inflammatory cytokine and chemokine receptors equip NMVs to competitively bind and neutralize excess proinflammatory mediators and curb abnormal immune overactivation. Meanwhile, intact adhesion biomarker repertoire sustains stable anchoring on inflamed vascular endothelium and achieves concentrated local immunoregulation at lesion sites [84,85]. Furthermore, the intact membrane framework of NMVs permits flexible surface peptide conjugation or therapeutic payload encapsulation for performance optimization. Owing to the essential regulatory status of neutrophils within autoimmune and inflammatory disorders, NMVs relying on their unique surface biomarker profiles serve as multifunctional biomimetic carriers for targeted inflammatory intervention [86]. Red blood cell-derived vesicles Red blood cell membrane-derived vesicles (RBC-MVs) are characterized by the expression of CD47, a “don't eat me” signal that prevents phagocytic clearance and confers prolonged circulation time in vivo [55,80]. Notably, RBC-MVs lack MHC molecules, which minimize their immunogenicity and make them ideal carriers for systemic drug delivery. Their unique immunological inertness and biocompatibility enable them to function as autoantibody decoys that sequester circulating pathogenic antibodies without triggering complement activation [55]. Engineered hybrid systems can be constructed by fusing erythrocyte membrane vesicles (RMVs) with synthetic carriers to combine the immune-evasive properties of RBC membranes with high drug-loading capacity [87]. Such hybrid vesicles retain native RBC membrane components and preserve key biological functions, supporting efficient cargo delivery and modulation of antigen-presenting cell activities. Similarly, RBC membrane-coated polymeric nanoparticles optimize pharmacokinetic profiles and enable sustained drug release. Owing to the immune-evasive properties of RBC membranes, these particles achieve an in vivo circulation half-life of approximately 40 h, along with improved delivery precision and biosafety [88,89]. Moreover, intact RBC membrane-cloaked nanoparticles can preserve native antigenic profiles to specifically sequester targeted autoantibodies, functioning as membrane-based antibody traps without inducing broad-spectrum immunosuppression [90]. Collectively, RBC-MVs provide a versatile and non-immunogenic platform for both immunosuppressive drug delivery and antibody sequestration. Platelet-derived vesicles Platelet-derived membrane vesicles (PLT-MVs) and platelet membrane-coated nanoparticles (PNPs) have emerged as promising biomimetic platforms due to their inherent affinity for inflamed or thrombogenic vascular sites. This inflammatory tropism arises from the presence of key platelet surface proteins-such as GPIb-IX-V, αIIbβ3, and P-selectin-which mediate adhesion to activated endothelium and exposed subendothelial components [91]. PNPs are typically constructed with biodegradable PLGA cores camouflaged by native platelet membranes, which fully retain intrinsic platelet surface antigens and biological characteristics [92]. Relying on complete membrane protein inheritance, these biomimetic nanoparticles can serve as functional decoys for pathogenic immunoglobulins, enabling targeted antibody trapping without interfering with normal hemostatic function or inducing broad systemic immunosuppression [92]. By integrating platelet membranes onto polymeric nanoparticle cores, versatile dual-targeting biomimetic platforms can be fabricated. The preserved P-selectin and glycoprotein VI (GPVI) on the membrane enable selective recognition and binding to inflammation-related adhesion molecules and matrix proteins, endowing the vesicles with excellent inflammatory tissue tropism [93]. Combined with controllable drug loading capability, platelet membrane coating effectively enhances the accumulation and retention of therapeutic cargo at inflammatory sites, improving the precision of localized immune regulation [93]. Furthermore, platelet-derived vesicles possess inherent vascular adhesive properties and biological regulatory functions. They can modulate inflammatory cell infiltration, regulate the expression of pro-inflammatory cytokines, and interfere with inflammation-related signaling cascades, thereby remodeling the pathological inflammatory microenvironment [94]. Altogether, platelet-derived vesicles offer a versatile and immune-compatible platform for targeted immune modulation. Their natural homing capabilities, intact surface protein profiles, and customizable drug-loading properties render them a reliable and precision-oriented biomimetic nanoplatform for diverse inflammatory and immune-related pathological conditions. Hybrid vesicles Hybrid vesicles, formed by combining membranes from multiple immune cell types, integrate the unique properties of their parental cells to enhance homing capability, functional versatility, and therapeutic efficacy [95]. By leveraging complementary surface proteins and biological functions, these vesicles overcome the limitations of single-cell-derived vesicles and broaden the potential for immune modulation in autoimmune disorders [96]. Various hybrid vesicle systems have been developed to integrate dual or multiple cellular advantages. Neutrophil-macrophage hybrid vesicles integrate the acute inflammation-targeting adhesion molecules of neutrophils and the immunoregulatory receptors of macrophages, enabling dual advantages of active inflammatory targeting and cytokine regulatory capability [97]. Such multi-membrane integration endows vesicles with multi-dimensional immunomodulatory potential, superior to single-cell membrane platforms. Erythrocyte-liposome hybrid vesicles are constructed by fusing natural RMVs with synthetic liposomal structures [87]. These hybrids retain erythrocyte-derived membrane components such as CD47 for immune evasion and selective cellular interaction, while inheriting the high drug-loading capacity and structural tunability of artificial liposomes. Through standardized co-extrusion fabrication, the hybrid structure maintains stable particle size and uniform potential, successfully integrating the biological advantages of natural cell membranes and the delivery advantages of synthetic carriers [87]. Erythrocyte-platelet hybrid membrane nanoparticles are fabricated via the fusion of RBC and platelet membranes at controlled protein ratios, with successful membrane interspersion verified by Förster resonance energy transfer (FRET) detection [98]. The resulting nanoparticles retain characteristic surface markers of both cell types, combining RBC-derived CD47-mediated immune evasion and platelet-derived inflammatory targeting ability. This hybrid structure achieves an in vivo elimination half-life of 51.8 h, while maintaining strong tissue-binding capacity, thus realizing optimized delivery performance [98]. Macrophage-T cell hybrid membrane nanoparticles integrate characteristic surface markers from both macrophages and T cells [99,100]. The dual membrane profile enables selective recognition and homing to inflammatory immune cells via distinct chemokine axis interactions. Combined with functional inner cores, these hybrid vesicles achieve efficient intracellular regulation of immune cell pyroptosis and inflammatory signaling, realizing effective modulation of complex inflammatory microenvironments [99,100]. Overall, hybrid vesicles serve as a highly modular and customizable biomimetic platform. Different parental cell membranes can be flexibly selected and fused according to the characteristics of inflammatory microenvironments, to synergize targeting performance, immune signaling regulation, and microenvironmental remodeling functions. This modular design makes hybrid vesicles a versatile and advanced tool for comprehensive immune modulation. Table 2, Table 3 collectively offer a full profile of IACMVs. Table 2 compares key quantitative parameters and targeting mechanisms across all vesicle types, while Table 3 describes their functional characteristics, pros and cons, in vivo performance and applicable autoimmune diseases. Table 2Quantitative performance comparison of IACMVs.Table 2Type of IACMVsCirculation Half-LifeTarget Tissue EnrichmentCytokine Neutralization IC50/Binding CapacityIn Vivo ToxicityReferenceMMVsMM-NPs exhibited an elimination half-life of 17.2 h in mice after intravenous injectionThese vesicles migrate along chemokine gradients and via VLA-4/VCAM-1 interactions. They display general inflammation tropism but possess low selectivity for diseased tissues.Binding Capacity:TNF-α: 4.3 pg/mgIL-6: 105.1 pg/mgIFN-γ: 6.5 pg/mgLow[101,102]DCMVsDCMVs exhibited short circulation (half-life not quantified) with rapid homing to lymph nodes and spleenSpecifically targets T cells and secondary lymphoid organs via MHC-I/II and CD80/86Not reportedLow[81,103]NMVsNM-NPs exhibited a blood circulation half-life of 2.017 ± 0.087 h following intravenous administrationVia membrane-anchored LFA-1/Mac-1 integrins, as well as integrin β2–ICAM-1 and PSGL-1–selectin interactions, the vesicles firmly adhere to inflamed endothelium and target acute inflamed vasculature and ischemia–reperfusion lesions effectivelyIC50:IL-1β: 188 μg/mLTNF-α: 1,327 μg/mLLow[84,104]RBC-MVsRBC-NPs exhibited an elimination half-life of approximately 40 h in mice after intravenous administrationThese vesicles lack endogenous homing receptors and undergo non-specific passive targeting via an EPR-like effect. Their tissue accumulation depends on prolonged blood circulation instead of active homing to lesionsNot reportedLow[88,89]PLT-MVsPLT-MV conjugation extended aPDL1 half-life by 6.7-fold (5.2 → 34.8 h)Equipped with GPIb-IX-V, αIIbβ3 and other adhesion receptors, the vesicles adhere to exposed collagen and damaged endothelium, and actively home to injured vessels, thrombi and inflammatory sites for local drug deliveryNot reportedLow[50,105]Hybrid VesiclesRBC-platelet hybrid vesicles achieved a half-life of 51.8 hIntegrates targeting modalities from source membranes; e.g., macrophage-T cell hybrid nanoparticles target inflamed myocardium via CXCL9/10/11-CXCR3 axis and integrin-mediated adhesionNot reportedLow[98,106] Quantitative performance comparison of IACMVs. Table 3Head-to-head comparison of diverse IACMVs: key properties and therapeutic applicability for autoimmune diseases.Table 3Cell Membrane Vesicles TypeCore Characteristics & AdvantagesLimitationsIn vivo Behavioral CharacteristicsPreferred Autoimmune Disease ScenariosReferencesMMVsPreserve characteristic membrane proteins (integrin α4β1, CD14, scavenger receptors, CD47) and are categorized into proinflammatory M1/tissue-repairing M2 subtypes; enable inflamed endothelium targeting, immune evasion and inflammatory cytokine sequestration, with wide application potential for chronic inflammatory interventionLarge batch heterogeneity determined by macrophage polarization status; M1-derived vesicles carry proinflammatory cargos with inflammation exacerbation risk; polarization culture leads to low preparation yield and poor scalabilityModerate blood circulation; preferentially accumulate at inflamed vessels; readily internalized by local immune cells and cleared via reticuloendothelial systemChronic persistent inflammatory autoimmune disorders: ulcerative colitis, chronic RA, diabetic immune-related complications[47,59,77,78,[107], [108], [109]]DCMVsPossess intact MHC-I/II, costimulatory molecules (CD80/CD86) and adhesion proteins; nitrogen cavitation produces homogeneous vesicles with complete membrane protein retention; enable flexible regulation of immune activation/tolerance, with uniform pharmacokinetic performance and feasibility for drug/nucleic acid loading to induce Treg differentiationStrict technical control required to avoid undesired immunostimulation; complicated modification for directional tolerance induction; high production cost and strict quality control barriers for clinical transformationPredominantly distributed in lymphoid organs and inflammatory lesions; weak long-distance circulation; mainly interact with adaptive immune cellsAntigen-driven autoimmune diseases requiring immune tolerance reconstruction: SLE,MS[49,80,81,[110], [111], [112]]NMVsAbundant surface LFA-1 (CD11a/CD18) binds endothelial ICAM-1 for acute inflammation tropism; embedded proteases, chemokine receptors and ROS-associated enzymes regulate inflammatory cascades; nitrogen cavitation yields uniform-sized vesicles with low PDI and favorable biocompatibility, with native active targeting, diversified inflammation modulation capacity and low immunogenicityPredominantly target broad inflammatory signals lacking pathway specificity; short in vivo half-life due to rapid clearance; limited durability for long-term chronic inflammatory treatmentRapid recruitment to acute inflamed vasculature; eliminated within minutes to hours by local immune cells and reticuloendothelial systemAcute-flare autoimmune diseases dominated by explosive vascular inflammation: active RA, autoimmune vasculitis[52,60,[84], [85], [86],[113], [114], [115], [116], [117]]RBC-MVsExpress surface CD47 for self-recognition and immune escape; absent MHC antigens to reduce inherent immunogenicity, suitable as soluble factor decoy carriers; with excellent long-circulation property, superior biosafety and minimal intrinsic immunogenicity, ideal for systemic intravenous administrationNo endogenous active targeting capacity and rely solely on passive EPR accumulation; restricted intrinsic immunomodulatory capacity and limited loading efficiency for macromolecular cargoProlonged blood retention with low cellular uptake; lesion deposition mostly depends on passive permeationAutoantibody-mediated systemic autoimmune pathologies: AIHA, SLE with lupus nephritis[55,87,88,90,[118], [119], [120]]PLT-MVsSurface GPIb-IX-V, αIIbβ3 and P-selectin support specific adhesion to injured inflamed vasculature; with natural tropism for damaged thrombogenic endothelium, available as neutralizing decoy against autoantibodiesRetained thrombogenic components raise thrombosis risk; limited raw material supply restricts large-scale manufacture; inherent hemostatic function interferes with in vivo therapeutic outcomesSelectively enrich at vascular injury sites; moderate circulation lifespan, dynamically interact with circulating platelets and vascular endothelial cellsVascular-injury-dominated autoimmune disorders: ITP, SLE vasculitis, IBD[[91], [92], [93], [94],[121], [122], [123]]Hybrid VesiclesConstructed by fusion of multiple cell plasma membranes to inherit complementary biofunctions from source cells; integrate multi-source membrane advantages to upgrade targeting specificity and multifunctionality, compensate inherent defects of single-cell-derived vesiclesUnstable membrane fusion ratio leads to poor batch reproducibility; difficult quantitative detection of membrane protein orientation and activity; multi-cell cultivation markedly elevates manufacturing costBiological behavior varies with component proportion of fused membranes, combining the biodistribution traits of all parent cell sources to fit complicated inflammatory microenvironmentComplex, multifactorial autoimmune inflammation with overlapping immune pathological pathways, multi-system involved autoimmune diseases[87,[97], [98], [99], [100],[124], [125], [126]] Head-to-head comparison of diverse IACMVs: key properties and therapeutic applicability for autoimmune diseases. Preparation, functionalization and engineering of IACMVs Membrane isolation techniques Non-nucleated cells (e.g., erythrocytes, platelets) undergo membrane isolation through density gradient centrifugation for cell separation, followed by hypotonic lysis or freeze-thaw cycles for disruption. Subsequent centrifugation removes soluble proteins, and extrusion or sonication is used to form nanovesicles. Nucleated cells (e.g., macrophages, tumor cells, stem cells) undergo similar procedures, employing mechanical homogenization or ultrasonication for disruption, and gradient or differential centrifugation to eliminate nuclei and cytoplasmic proteins [[127], [128], [129], [130], [131]] (Fig. 2A). Schematic illustration of method of preparation, functionalization, and characterizations of IACMVs (A) Method of preparation of IACMVs, involving isolation of cell membrane, core nanoparticles and coating of cell membrane on nanoparticles. (B) Various techniques used for functionalization of IACMVs. (C) Characterization of IACMVs. All schematics in this figure are original and self-drawn. Hypotonic lysis exploits osmotic pressure differences to rupture cells by exposing them to low-osmolarity solutions, which causes water influx, swelling, and membrane rupture [132]. This gentle technique minimizes damage to membrane proteins. For example, RBC membranes used in SLE therapy are isolated via hypotonic lysis, followed by centrifugation to remove hemoglobin and debris, thereby preserving CD47-mediated immune evasion [88]. Neutrophil membranes for colitis treatment retain adhesion receptors such as CD11b when isolated using hypotonic buffers [86], while platelet membranes obtained by this method preserve integrins (e.g., GPIIb/IIIa) that are critical for targeting inflamed endothelium [94]. B cell membranes used in multiple myeloma therapy rely on hypotonic lysis to preserve BAFF/APRIL receptors [133]. This method is ideal for fragile cells-such as erythrocytes, platelets, and neutrophils-where mechanical disruption could denature membrane proteins [50,86,88,133]. Homogenization employs mechanical forces (e.g., rotor-stator homogenizers) to shear cell membranes. For atherosclerosis therapy, macrophage membranes are isolated by homogenizing RAW 264.7 cells on ice, preserving CCR2 and CD36 to target inflamed plaques [47]. This homogenization process ensures the retention of functional membrane components, which are crucial for the ability of MSC-derived vesicles including nanovesicles and synthetic eukaryotic vesicles to attenuate pro-inflammatory cytokine release and modulate immune responses [44]. Furthermore, mechanical homogenization facilitates the isolation of plasma membrane fractions from the genetically engineered leukemia cell line K562, which is modified to express bispecific antibodies, costimulatory ligands, cytokines, and immune checkpoint-blocking antibodies. These membrane fractions can be reconstituted into membrane-coated nanoparticles or ultrasonicated into nanovesicles for evaluating antitumor functions both in vitro and in vivo [134]. Sonication employs high-frequency sound waves (40-44 kHz) to disrupt membranes through cavitation and is often combined with centrifugation for purification. MSC-derived synthetic vesicles (SyEVs) designed for systemic inflammation are prepared by sonicating cell lysates, which removes cytosolic components while preserving membrane-bound immunosuppressive factors [44]. Neutrophil membranes used in RA treatment undergo sonication during the preparation of peptide-anchored neutrophil membrane-coated biomimetic nanodrugs (R4F-NM@F127). This step, integral to membrane isolation and nanocarrier construction, facilitates the formation of uniform structures. The neutrophil membrane retains CXCR2, enabling the nanodrug to be recruited to inflamed joints via the chemotactic effect of IL-8, which is highly expressed in RA synovial fluid, thereby enhancing targeting to inflamed joint tissues [135]. Similarly, in the preparation of neutrophil (N)-macrophage (M) hybrid membrane-coated PB (NM-PB) for ulcerative colitis, sonication at 37 °C for 10 min fuses neutrophil and macrophage membranes at a 1:1 protein ratio. Subsequent sonication in an ice bath for 10 min coats the hybrid membrane onto PB nanozymes. This process preserves membrane proteins that facilitate targeting to inflamed colon tissues, thereby improving therapeutic efficacy [97]. Freeze-thaw cycles were employed to derive platelet membranes. Platelets isolated from mouse whole blood were repeatedly frozen at −80 °C, thawed at room temperature, and centrifuged at 4000 × g for 3 min, undergoing at least three cycles to extract the membranes. These membranes, which retained key functional proteins such as P-selectin and integrin β3, were coated onto PA-loaded PLGA-PEG nanoparticle cores via sonication to form PNPs [94,136]. The resulting PNPs demonstrated targeted accumulation in inflamed colon tissue, alleviated DSS-induced acute colitis in mice, and mitigated endothelial inflammation by modulating the Ca2+-ROS signaling pathway [94]. Similarly, in the preparation of neutrophil membrane-coated PSA nanoparticles (N-PSAs) for rheumatoid arthritis treatment, freeze-thawing was used to derive membranes from frozen neutrophils. These cells were thawed, washed, lysed, and centrifuged to obtain membranes retaining proteins such as LFA-1 and Ly-6G. The membranes were then coated onto PSA nanoparticles via sonication. N-PSAs targeted inflamed joints and released salicylic acid to alleviate symptoms by inhibiting NETs through the MAPK signaling pathway [113]. Extrusion is a critical step in preparing leukocyte membrane-cloaked nanoparticles (Leu-Mem-NPs). After isolating leukocytes from whole blood via centrifugation and hypotonic lysis, the extracted membranes undergo extrusion, combined with sonication, to form spherical vesicles. This process, together with ultrasonication, is also employed to fuse membrane vesicles with core nanoparticles, utilizing tensile stress to promote fusion and form core-shell structures while preserving membrane integrity and surface proteins [52]. In the preparation of berberine-loaded macrophage membrane-derived nanovesicles (Ber@MVs) for ulcerative colitis treatment, extrusion plays a vital role. Specifically, the isolated macrophage membranes and berberine are dispersed in PBS and sequentially passed through 1 μm, 400 nm, and 200 nm polycarbonate membrane filters via 15 consecutive extrusion to form Ber@MVs. This process not only encapsulates berberine within the membrane vesicles but also ensures uniform vesicle size and preserves membrane proteins, which are essential for targeting inflammatory sites and exerting anti-inflammatory effects [78]. Extrusion is also crucial in fabricating hybrid R-Lipo for systemic lupus erythematosus therapy. The process involves first extruding erythrocyte membranes through 400 nm polycarbonate membranes using a mini-extruder to form RMVs. These RMVs are then co-extruded with MPA-R-Lipo [87]. This method ensures the successful fusion of natural erythrocyte membranes with artificial liposomes, preserving membrane components and enabling targeted delivery to dendritic cells, which is key for inhibiting their maturation and alleviating lupus nephritis [87]. Vesicle fabrication and coating methods The fabrication of IACMVs primarily involves fusing isolated cell membranes with nanoparticle cores to form core-shell structured vesicles (Fig. 2A.). Various techniques offer distinct trade-offs in scalability, precision, and protein retention [46]. Extrusion is widely used to pass membrane-nanoparticle mixtures through polycarbonate membranes. This method excels at producing uniform vesicles with well-preserved membrane protein profiles and functional receptor-ligand interactions, which are critical for biological activity and targeting [58,133]. It reliably maintains antigen orientation and membrane integrity, as demonstrated in macrophage or neutrophil membrane coatings [52,57]. However, multiple passes may be required to achieve homogeneity, and the process can impose mechanical stress on delicate membrane components, potentially affecting labile proteins. Sonication is a scalable technique that employs ultrasonic waves to facilitate membrane-nanoparticle fusion, making it suitable for coating PLGA or mesoporous silica nanoparticles [44,80]. This method preserves membrane ligand functionality and is relatively easy to implement for large-scale production. However, its drawbacks include potential localized heating, which may denature heat-sensitive proteins, and an increased risk of vesicle size polydispersity compared to extrusion. Nitrogen cavitation facilitates gentle vesicle formation under controlled pressure, minimizing physical damage to membranes. This method ensures high protein retention and batch reproducibility, making it ideal for sensitive cell types such as dendritic cells and neutrophils [49,60]. However, it is less scalable than extrusion or sonication and requires specialized equipment, which limits its widespread use. Microfluidic electroporation provides fine control over fusion parameters through the application of electrical pulses, enabling continuous-flow fabrication and high reproducibility [57]. This fineness of control makes the technique well-suitable for large-scale applications and complex membrane engineering. However, its main limitations include the requirement for advanced microfluidic setups and the potential disruption of membrane integrity if electrical parameters are not properly optimized. Chemically induced vesiculation relies on chemical stimulants to trigger membrane budding, which can be applied alone or coupled with mechanical extrusion for the fabrication of nano plasma membrane vesicles (nPMVs) [40]. This method enables fine control over vesicle properties but may leave residual chemicals that compromise biocompatibility. Ultrasonic reassembly, a related technique applied to vesicles derived from genetically modified cells, preserves targeting ligands; however, it carries the risk of protein misfolding due to mechanical stress [137]. Across all methods, the primary challenge is balancing efficiency with the preservation of native membrane structure, protein orientation, and biological activity-factors critical to therapeutic functionality of IACMVs [46,49,58,138] (refer to Table 4). Table 4Techniques for preparation, fabrication, and engineering of IACMVs.Table 4Technique CategorySpecific TechniqueAdvantagesDisadvantagesReferencesMembrane Isolation TechniquesHypotonic LysisGentle technique minimizing damage to membrane proteinsOnly effective for osmotic pressure-sensitive cells[50,86,88,133]Ideal for fragile cells (erythrocytes, platelets, neutrophils)Requires additional centrifugation to remove soluble proteins and debrisPreserves functional receptors (CD47 on RBCs, CD11b on neutrophils, GPIIb/IIIa on platelets, BAFF/APRIL receptors on B cells)HomogenizationPreserves functional membrane components (CCR2 and CD36 on macrophages, immunosuppressive factors on MSCs)Mechanical force may denature labile proteins (requires temperature control)[44,47,134]Suitable for cells resistant to hypotonic lysisRisk of cytoplasmic contamination requiring purificationEnables isolation from genetically engineered cellsSonicationPreserves membrane-bound immunosuppressive factors in MSC-derived synthetic vesiclesLocalized heating may damage heat-sensitive proteins (requires cooling)[44,97,113,135]Retains neutrophil membrane proteins (CXCR2) for joint targeting in RAPotential for vesicle size polydispersity if parameters unoptimizedFacilitates hybrid membrane fusion (e.g., neutrophil-macrophage at 1:1 protein ratio)Freeze-Thaw CyclesPreserves key functional proteins (P-selectin, integrin β3 on platelets; LFA-1, Ly-6G on neutrophils)Multiple cycles prolong processing time[94,113,136]Compatible with downstream coating processesRepeated thermal stress may cause partial protein denaturationWorks with frozen cell stocksExtrusionProduces uniform spherical vesicles with preserved membrane integrityMultiple passes increase mechanical stress on delicate proteins[52,78,87]Maintains surface proteins critical for targeting and therapyParameter misoptimization may damage labile receptorsEnables membrane-nanoparticle fusion for core-shell structuresVesicle Fabrication & Coating MethodsExtrusionGenerates uniform vesicles with well-preserved membrane protein profilesMechanical stress may affect labile membrane components[52,57,58,133]Maintains functional receptor-ligand interactionsRequires strict control of filter pore sizes for homogeneityReliably maintains antigen orientation and membrane integritySonicationScalable technique suitable for large-scale productionLocalized heating may denature heat-sensitive proteins[44,80,97]Preserves membrane ligand functionalityHigher size polydispersity compared to extrusionFacilitates hybrid membrane coatingNitrogen CavitationGentle process minimizing physical membrane damageLess scalable than extrusion or sonication[49,60]Ensures high protein retention and batch reproducibilityRequires specialized equipmentIdeal for sensitive cell types (dendritic cells, neutrophils)Microfluidic ElectroporationFine control over fusion parameters for consistent productionRequires advanced microfluidic setups[57]Enables continuous-flow fabrication for large-scale applicationsElectrical parameter misoptimization may disrupt membrane integrityPreserves native protein orientation for enhanced targetingChemically Induced VesiculationEnables fine control over vesicle properties (size, protein profile)Residual chemicals may compromise biocompatibility[40,53,137]Compatible with extrusion to improve monodispersityIn SLE, apoptotic PMVs may exacerbate inflammation via cGAS-STINGCan induce apoptotic features for immunomodulationFunctionalization & Engineering TechniquesGenetic EngineeringStably incorporates target proteins/receptors into vesiclesRequires viral vector expertise and cell transfection[45,133,139]Enables specific immune signaling modulationRisk of off-target expression or integration artifactsPreserves protein function through native expressionBiochemical EngineeringMild reactions preserve vesicle integrity and functionalityLimited to molecules that conjugate to native membrane components[44]Enhances natural membrane properties without residual chemicalsLower loading efficiency compared to other methodsMaintains low immunogenicity for autoimmune therapyChemical EngineeringGenerates uniform vesicles with engineered surface profilesResidual chemicals may reduce biocompatibility[40,137]Scalable for large batch productionUltrasonic reassembly risks protein misfoldingEnables small molecule/cytokine incorporationPhysical EngineeringPreserves native protein orientation for targeting efficiencyRequires specialized equipment[57,60,136]Enables hybrid vesicle fabrication without chemical modifiersSensitive to electrical parameter optimizationReduces toxicity concernsMembrane HybridizationSynergizes advantages of multiple parental membranesComplex fusion protocols requiring fine control[87,[97], [98], [99], [100],140]Enhances functionality and targeting capabilitiesUneven protein distribution may reduce efficiencyEnables multimodal therapeutic approaches Techniques for preparation, fabrication, and engineering of IACMVs. Functionalization and engineering of vesicles Genetic engineering, chemical engineering, physical engineering and membrane hybridization are most common techniques used for functionalization of vesicles (Fig. 2B.). Genetic engineering is a powerful tool for enhancing the functionality of IACMVs by modifying source cells to express specific proteins or receptors, which are stably retained in vesicles during fabrication. For example, Fang et al. used lentiviral transduction to engineer fibroblasts to express CD40, a key co-stimulatory molecule. Nanovesicles derived from these cells via extrusion and sonication maintained surface CD40 and were co-loaded with mycophenolate mofetil (MMF), enabling modulation of T-B cell crosstalk in SLE models [45]. Similarly, He et al. genetically modified HEK-293T cells to express recombinant BCMA, and vesicles fabricated through extrusion displayed BCMA on their surface, allowing selective sequestration of pathogenic BAFF and APRIL ligands involved in plasma cell survival [133]. Wang et al. further extended this strategy by engineering CHO-K1 cells to express a fusion protein linking anti-CD8 scFv and IL-2 mutein via a granzyme B-cleavable linker, constructing bispecific nanovesicles (aCD8-GrzBcs-IL2 MVs) that selectively depleted cytotoxic CD8+ T cells while promoting Treg expansion in type 1 diabetes models [139]. Biochemical engineering involves conjugating bioactive molecules to native membrane components through mild reactions that preserve both vesicle integrity and molecular functionality. Park et al. extracted membranes from mesenchymal stem cells and functionalized them with Myd88-targeting peptides via mild alkaline incubation. Subsequent sonication produced SyEVs that retained cytokine-responsive properties, enabling specific inhibition of TLR4-Myd88 signaling in inflammatory cells [44]. These SyEVs, valued for their low immunogenicity and stability, exemplify how biochemical modification can enhance natural membrane functions for autoimmune therapy. Chemical engineering utilizes small molecules or chemically induced vesiculogenesis to modify membrane properties, often in combination with other techniques to enhance scalability. Alter et al. combined chemically induced vesiculogenesis-using reagents such as calcium ionophores-with mechanical extrusion to produce monodisperse nPMVs, preserving engineered surface profiles suitable for large-scale applications [137]. These nPMVs exhibit uniform sizes, enriched phosphatidylserine (PS) expression, and simplified functional protein profiles, effectively addressing the heterogeneity found in natural vesicles. Ren et al. employed ultrasonic fragmentation to fabricate bio-MVs from chemically engineered cells, ensuring the retention of cytokine-binding domains for potent immunomodulatory activity [40]. Additionally, PMVs generated during apoptosis or chemical induction retain apoptotic features such as PS exposure, which can engage phagocyte receptors to promote immune resolution. However, in SLE, apoptotic lymphocyte-derived PMVs may exacerbate inflammation through dsDNA-mediated activation of the cGAS-STING pathway [53,54]. Physical engineering utilizes mechanical or electrical forces to integrate functional ligands into membranes while preserving the native orientation of proteins. Qiu et al. employed microfluidic electroporation to accurately incorporate ligands into membranes during vesicle fabrication, thereby enhancing targeting accuracy by ensuring correct ligand orientation [57]. Gao et al. demonstrated that physical extrusion of LPS-stimulated neutrophils preserved β2 integrin expression on NMVs, which maintained an optimal size distribution and adhesion to activated platelets, validating their potential for targeted delivery to inflammatory sites [60]. Beyond natural membrane sources, synthetic hybrid vesicles exemplify the versatility of physical engineering. For instance, Guo et al.’ s IM-MNPs/dexamethasone (DXM) vesicles coat dexamethasone-loaded liposomes with membranes derived from IFN-γ stimulated 4T1 cells, inheriting surface proteins such as PD-L1, CD155, CD47, and CD24. These vesicles effectively target pathogenic CD4+ T cells, suppress effector responses, and promote regulatory T cell differentiation in lupus models [136]. Membrane hybridization, the core strategy for constructing hybrid vesicles, involves fusing membrane components from multiple distinct sources (e.g., different cell types, natural cell membranes with synthetic liposomes) to integrate the unique properties of parental membranes into a single vesicle system [140]. This approach enables hybrid vesicles to overcome the functional limitations of single-source vesicles by synergizing complementary advantages: for instance, neutrophil-macrophage hybrid membranes combine the acute inflammation-targeting ability of neutrophils (via LFA-1) with the immunoregulatory functions of macrophages (scavenger receptors for cytokine sequestration, M2 polarization induction) to treat IBD [97]; erythrocyte-liposome hybrids (R-Lipo) leverage erythrocyte CD47-mediated DC recognition for targeting and liposomal efficient drug loading to deliver immunosuppressants for SLE [87]; erythrocyte-platelet hybrids ([RBC-P] NPs) integrate RBC's CD47-driven immune evasion and platelet-mediated targeting for atherosclerotic plaque accumulation; and macrophage-T cell hybrids (siIRF1@ZIF@HM NPs) utilize macrophage ITGA4/ITGB1 and T cell TCR-β to actively target inflamed myocardium and suppress macrophage pyroptosis in autoimmune myocarditis [99,100]. Besides disease-specific targeting, hybrid IACMVs also support multimodal therapeutics: Luo et al. developed M2 macrophage membrane-coated Janus nanomotors with MnO2 cores, enabling ROS-responsive propulsion and cytokine scavenging for colitis therapy [59]; Han et al. used platelet-mimetic vesicles for siRNA delivery and photothermal therapy to achieve controlled release for autoimmune inflammation intervention [91]; Harvey et al. showed DC-derived nanovesicles (via nitrogen cavitation) retain CD80/CD86/MHC to support T cell activation [49]; and Mohale et al. highlighted leukocyte-based hybrid membrane nanoparticles that target inflamed tissues (e.g., RA), scavenge pro-inflammatory cytokines, and promote M2 polarization [52]. Collectively, these strategies-spanning genetic modification, biochemical conjugation, chemical induction, and physical engineering-highlight the adaptability of engineered and synthetic vesicles. By bridging natural immune mimicry with synthetic precision, they enable functionalization with immune checkpoints, cytokine decoys, or antigenic peptides, positioning IACMVs as a transformative platform for precision immunotherapy in autoimmune disorders [44,45,133]. Characterization and quality control of IACMVs Physical and morphological characterization IACMVs require a complementary combination of techniques to quantify key parameters and visualize structural features, ensuring functional integrity and batch consistency (Fig. 2C.). DLS and nanoparticle tracking analysis (NTA) are primary tools for evaluating size distribution and concentration. For example, red blood cell membrane (RBCM)-coated nanoparticles (Hb-Naf@RBCM NPs) exhibit a hydrodynamic diameter of approximately 62 nm, while uncoated cores measured around 45 nm, indicating successful membrane coating [141]. Neutrophil membrane-encapsulated nanoparticles show a mean diameter of approximately 150 nm via DLS [85], and dendritic cell membrane-derived nanovesicles display a narrow size range of 60–300 nm, with an average diameter of 123-125 nm as measured by NTA [49]. These measurements are consistent with reported IACMVs sizes ranging from 100 to 200 nm, which is critical for biological trafficking and targeting [58,133,138]. Morphological analysis using TEM and cryo-EM confirms structural uniformity. Hb-Naf@RBCM NPs exhibit a distinct 7 nm membrane coating [141], while neutrophil membrane-coated PLGA nanoparticles display a well-defined bilayer and spherical morphology [85]. Cell-derived nanovesicles visualized by field-emission scanning electron microscopy (FESEM), appear predominantly spherical, with occasional irregularities attributed to sample preparation [49]. PNPs demonstrate a unilamellar structure under TEM [93]. Surface charge, measured by zeta potential, is influenced by anionic membrane components and typically ranging from −20 to −30 mV. Macrophage membrane-camouflaged liposomes exhibit a zeta potential of −28.0 mV [77], while platelet membrane-coated nanoparticles display approximately −24.7 mV [93]. Engineered PD-1/TIGIT dual-activating cell membrane nanoparticles (IM-MNPs/DXM) have a less negative zeta potential due to surface shielding by membrane proteins, which enhances colloidal stability and reduces nonspecific adsorption of serum proteins [136]. Biochemical characterization and membrane protein retention Biochemical validation confirms the retention of both native and engineered proteins, as well as the integrity of cargo. SDS-PAGE and Western blotting are serve as foundational techniques: Sun et al. and He et al. confirmed the retention of CD47, integrins, and recombinant BCMA on macrophage-derived and engineered vesicles [58,133]; dendritic cell-derived vesicles preserve MHC-I, CD80, and CD86 [49]; and neutrophil membrane vesicles retain β2 integrins and cytokine receptors, such as TNF-αR and IL-1βR [93,114]. These proteins are critical for targeting, adhesion, and immune modulation. Beyond basic detection, assays are used to validate functional integrity. Park et al. confirmed that MyD88 inhibitors in SyEVs retain the ability to disrupt inflammatory protein-protein interactions by using Western blotting combined with functional readouts [44]. Kato et al. detected genomic dsDNA in apoptotic PMVs through qPCR and DNase I resistance assays, linking nucleic acid retention to immunostimulatory potential [53]. Li et al. validated the site-specific conjugation of biotinylated peptides on vesicle surfaces, ensured that engineered ligands maintained their targeting capability [80]. He et al. further verified the functionality of recombinant BCMA through ligand-binding assays, demonstrating specific interaction with BAFF/APRIL to block pathogenic signaling in autoimmune models [133]. Functional assays and immunological profiling Functional validation confirms that IACMVs retain therapeutic activity relevant to autoimmune disease outcomes. Mohale et al. demonstrated the retention of integrins and CD47, ensuring targeted delivery to disease sites via macrophage and neutrophil vesicles [52]. Baffour et al. employed Annexin V-FITC staining and flow cytometry to verify phosphatidylserine exposure on apoptotic PMVs, indicating their immunomodulatory potential [54]. Gao et al. showed that neutrophil-derived NMVs retain β2 integrins, facilitating adhesion to activated platelets and confirming their targeting capability [60]. Efficacy is confirmed through immunological assays. Jiang et al. demonstrated that immune membrane-derived nanosponges scavenge circulating autoantibodies, thereby reducing pathogenic titers through antigen decoy mechanisms [55]. Harvey et al. employed antigen-specific T-cell activation assays-measuring CD69 expression and IFN-γ production to show that DC-derived vesicles retain MHC and co-stimulatory molecules, confirming their capacity to trigger adaptive immune responses [49]. Li et al. engineered vesicles to either suppress or activate the immune system, as evidenced by T-cell proliferation and cytokine profiling [80]. Sun et al. and Fang et al. validated the multi-step functionality of IACMVs, demonstrating that they evade phagocytosis via CD47-SIRPα signaling and selectively deliver payloads to inflamed tissues [45,58]. Stability, storage, and safety evaluation Stability and safety assessments are essential for clinical translation. Luo et al. demonstrated that M2 macrophage membrane-coated Janus nanomotor vesicles maintain stable size and zeta potential for eight days when stored at 4 °C and −20 °C, supporting long-term storage [59]. Gao et al. employed high-performance liquid chromatography to characterize the sustained release of payload from neutrophil-mimetic vesicles over 24 h, ensuring controlled dosing [60]. Alter et al. and Ren et al. confirmed batch-to-batch consistency in size, protein profiles, and efficacy for neutrophil-derived PMVs and Bio-MVs, thereby validating scalable manufacturing [40,137]. Safety evaluations include both in vitro and in vivo assays. Gao et al. reported low cytotoxicity of macrophage-derived vesicles, while Han et al. demonstrated minimal liver and kidney toxicity in animals treated with platelet-mimetic vesicles [47,91]. Collectively, these findings establish IACMVs as stable, safe, and reproducible platforms, supported by robust quality controls that endorse their application in autoimmune disease therapy [40,47,59,60,91,137]. Therapeutic applications of IACMVs in autoimmune diseases Inflammatory bowel disease Immune cell membrane-coated nanomedicines have demonstrated significant therapeutic potential in IBD, including ulcerative colitis and Crohn's disease, by harnessing natural targeting abilities and multifunctional mechanisms. Macrophage-derived vesicles and their engineered variants have been extensively investigated for IBD therapy. Sun et al. demonstrated that macrophage-derived vesicles naturally accumulate at inflamed intestinal sites due to inherent inflammatory tropism. These vesicles scavenge pro-inflammatory cytokines such as TNF-α and IL-1β and feature ROS-responsive cores that enable drug release in oxidative environments. This mechanism promotes the polarization of macrophages from the pro-inflammatory M1 phenotype to the anti-inflammatory M2 phenotype, facilitating inflammation resolution [58]. In DSS-induced colitis models, the Motor@M2M@ self-assembling matrix (SAM) formulation significantly reduced levels of TNF-α and IL-6, increased IL-10 expression, improved histological scores, and restored epithelial barrier integrity by upregulating tight junction proteins, including ZO-1 and occluding [59] (Fig. 3.). Ma et al. developed Ber@MVs-CA nanoparticles that utilize macrophage membranes for targeted delivery to inflamed tissues and a calcium alginate coating for pH-responsive release of berberine (87% release under colonic conditions). These nanoparticles scavenge pro-inflammatory cytokines, reprogram macrophages toward the M2 phenotype, reduce oxidative stress, and enhance mucosal healing [78]. Similarly, nanoparticles encapsulating rosiglitazone-a known inducer of polarization-within macrophage membranes target inflamed colon tissues, release the drug to promote M2 polarization, and absorb inflammatory mediators, showing potential in ulcerative colitis treatment [57,58]. Schematic Illustration of Motor@M2M@SAM Preparation and Its Therapeutic Mechanism in Ulcerative Colitis (UC) Management. (A) Diagram depicting the manufacturing process of Motor@M2M. (B) Diagram illustrating the production of Motor@M2M@SAM utilizing microfluidic technology. (C) Mechanism of UC Treatment: When administered orally, the SAM breaks down upon reaching the colon. Motor@M2M is subsequently released into the colonic lumen. The thrust from oxygen (O2) bubbles, generated by the decomposition of local hydrogen peroxide (H2O2) in the inflammatory microenvironment, promotes Motor@M2M's penetration through the mucus layer. These nanomotors target inflamed colon cells via macrophage-like properties, specifically binding to colon epithelial cells. Functioning as decoys, Motor@M2M scavenges inflammatory cytokines through receptor-ligand interactions and adsorption. Eventually, Motor@M2M treats UC by scavenging ROS, alleviating inflammation, reprogramming macrophages, repairing the epithelial barrier, and restoring the microbial balance in the intestine. This figure was reproduced with permission from Ref. [59]. Copyright 2022, AAAS. Neutrophil membrane-based systems have also demonstrated efficacy in IBD. Hu et al.‘s NM-PB targets inflamed colon tissues, scavenges ROS, suppresses cytokine production, promotes M2 macrophage polarization, and disrupts cytokine-cytokine receptor interactions, thereby providing dual anti-inflammatory and antioxidant effects for ulcerative colitis [97]. Neutrophil membrane-coated nanoparticles (MSN-HC@NM NPs) enable pH-responsive drug release (65% at pH 5.0) in DSS-induced colitis mice, reducing TNF-α and IL-1β levels and restoring tight junction proteins, resulting in superior histopathological improvement compared to uncoated nanoparticles [86]. Additionally, keratinocyte growth factor (KGF)-loaded neutrophil-like liposomes (KGF-Neus) enhance KGF stability and specifically target inflamed sites [142,143], significantly reducing intestinal mucosal ulcers and neutrophil infiltration while promoting recovery of intestinal morphologyin DSS-induced colitis models [144]. Other cell membrane-coated systems include leukocyte and platelet membrane formulations. Li et al. developed cyclosporine A-loaded nanoparticles coated with macrophage membranes (MM-CsA NPs) or leukocyte membranes (LM-CsA NPs), which utilize integrins α4β1 and α4β7 to target inflamed tissues by interacting with upregulated VCAM-1 and MadCAM-1 in IBD [145]. In vitro, these nanoparticles reduce phagocytosis by normal macrophages while increasing uptake by proinflammatory macrophages and endothelial cells. In vivo, they prolong retention in inflamed intestines; with LM-CsA NPs mitigate weight loss, colon shortening, and disease activity indices (DAIs) in both acute and chronic colitis, decrease levels of TNF-α, IL-6, and myeloperoxidase (MPO), and increase gut barrier proteins such as E-cadherin, ZO-1, and occludin [145]. Song et al.‘s PNPs deliver PA to inflamed vasculature by exploiting platelets' natural homing to injury sites, reducing pro-inflammatory cytokines and neutrophil infiltration, and restoring epithelial integrity through dual action on Ca2+-ROS signaling and barrier repair [94]. OX40L-blockade cellular nanovesicles (OX40 NVs), engineered to express OX40 receptors, accumulate in the inflamed colon through OX40-OX40L interactions. These nanovesicles reduce pathogenic CD4+OX40+ T cells, increase Tregs, decrease levels of TNF-α and IL-1β, alleviate weight loss, improve DAI, restore colon length, and reduce tissue damage, demonstrating superior efficacy compared to anti-OX40L antibodies [146]. Collectively, these cell membrane-coated nanomedicines enhance targeting precision, modulate immune responses, and repair mucosal barriers, providing multi-mechanistic strategies for the treatment of IBD. Systemic lupus erythematosus Immune cell membrane vesicle-based strategies offer promising therapeutic avenues for SLE, particularly by targeting dysregulated B cell and T cell responses. Although initially developed for multiple myeloma, recombinant BCMA-expressing nanovesicles (Re-BCMA-NVs) have demonstrated the ability to scavenge APRIL and BAFF-key mediators of B cell survival and differentiation-thereby suppressing NF-κB signaling, which is central to SLE pathogenesis [133]. In MRL/lpr lupus-prone mice, treatment with CD40-expressing nanovesicles (CD40 NVs), either alone or in combination with MMF, reduced anti-dsDNA and ANA autoantibodies, improved renal function, and limited immune complex deposition, highlighting their B cell-directed immunomodulatory potential [45]. Targeted delivery of immunosuppressants using biomimetic vesicles further enhances therapeutic precision and reduces systemic toxicity. Nanoparticles-hybrid systems, which combine erythrocyte membranes with liposomes, have been shown to effectively deliver MPA to DCs, suppressing their maturation and inflammatory cytokine production [87]. In MRL/lpr mice, MPA-R-Lipo treatment reduced CD40/CD80 expression and levels of TNF-α and IL-12p70, alongside a marked decrease in serum anti-dsDNA antibodies and renal pathology [87]. Similarly, red blood cell membrane-coated nanoparticles loaded with CsA-RNPs improved drug pharmacokinetics and bioavailability, significantly lowering inflammatory markers and renal damage in lupus models with minimal toxicity [88]. Beyond drug delivery, IACMVs also perform immunological functions, such as promoting immune tolerance and inhibiting pathogenic pathways. Neutrophil-derived PMVs, enriched with PS, mimic apoptotic bodies and induce tolerance through silent phagocytosis, offering therapeutic potential for SLE [137]. Reengineering pathogenic apoptotic vesicles, which are commonly elevated in SLE due to defective clearance, may enable them to act as decoys that scavenge autoantibodies or cytokines, thereby redirecting immune responses toward tolerance [54]. This approach could effectively modulate immune responses and promote tolerance in autoimmune diseases [54]. Furthermore, AdMVs activate the cGAS-STING pathway via their cytosolic DNA content, a key trigger of IFN-I signaling in SLE. Decoy vesicles lacking immunostimulatory DNA could competitively inhibit cGAS, thereby attenuating this pathogenic pathway [53]. Emerging biomimetic platforms are advancing immunomodulation strategies. IM-MNPs/DXM, a dexamethasone-loaded nanoparticle system coated with engineered 4T1 cancer cell membranes, achieves dual targeting of PD-1 and TIGIT on pathogenic CD4+ T cells [136]. This approach suppresses effector T cells while enhancing Treg function, thereby restoring immune balance. In MRL/lpr mice, IM-MNPs/DXM exhibited extended blood retention capability and specific targeting to diseased organs. The formulation effectively alleviates lupus nephritis with negligible systemic side effects [136]. (Fig. 4.). Schematic Illustration of DXM-Loaded IM-MNPs Preparation and Their Mechanism in SLE Treatment (A) Diagram outlining the preparation process of DXM-loaded IM-MNPs. Due to the lack of MHC I molecules, these MNPs exhibit low immunogenicity, which prevents allogeneic rejection and avoids unintended activation of the in vivo immune system. Additionally, treating 4T1 cells with IFN-γ leads to increased expression of CD155 and PD-L1 on the cell surface. (B) Schematic illustrating how IM-MNPs/DXM acts in SLE treatment: IM-MNPs/DXM carry two types of “don’ t eat me” receptors-CD47 and CD24-which work together to minimize the clearance of these nanoparticles by macrophages. Moreover, IM-MNPs/DXM can preferentially accumulate in inflamed organs: this targeting ability is achieved through the interaction between CD44 (an adhesion protein on the nanoparticle surface) and hyaluronic acid (HA), as HA is highly expressed at inflammatory sites. Once targeted, IM-MNPs/DXM bind to pathogenic CD4+ T cells and activate coinhibitory receptors; by interacting with PD-1 and TIGIT, they help regulate the balance between effector CD4+ T cells and Tregs. Meanwhile, DXM is released from IM-MNPs in a sustained manner: this released DXM inhibits the production of proinflammatory factors at the inflammatory site, further supporting Treg-mediated maintenance of immune homeostasis. This figure was reproduced with permission from Ref. [136]. Copyright 2022, Elsevier. Collectively, these studies highlight the diversity of vesicle-based therapies in addressing the multifactorial immune dysregulation characteristic of SLE, including B cell depletion, cytokine scavenging, T cell modulation, and tolerance induction. Rheumatoid arthritis Platelet membrane-coated vesicles have demonstrated significant therapeutic potential in RA by enabling targeted delivery of immunosuppressive agents. These vesicles employ adhesion molecules, including GPVI, to bind inflamed synovial tissues and endothelium through interactions with overexpressed CD44 and collagen IV [91,93]. In CIA models, FK506-loaded platelet vesicles markedly reduced joint inflammation, bone erosion, and local cytokine production, while enhancing nanoparticle retention and therapeutic concentration at disease sites. This cell-mimetic platform achieves favorable in vivo circulation and specific targeting to inflamed tissues, reducing systemic toxicity and boosting therapeutic outcomes [91,93]. Leukocyte membrane-coated nanoparticles show promising results in various inflammation-related disease models, including RA. Mohale et al. demonstrated that neutrophil membrane-cloaked nanoparticles preferentially accumulate in sites of inflammation, reduce levels of TNF-α and IL-6, and mitigate tissue damage [52]. These neutrophil membrane-coated nanoparticles represent a broad-spectrum, function-driven approach to RA therapy. They inherit immunoregulatory molecules such as LFA-1, IL-1R, and TNF-αR, enabling them to bind and scavenge key inflammatory cytokines like IL-1β and TNF-α [84]. In vitro, neutrophil-NPs suppress chondrocyte apoptosis, inhibit MMP-3 secretion, and promote aggrecan production. In vivo, they reduce joint swelling, leukocyte infiltration, and cartilage degradation, achieving effects comparable to anti-TNF-α or anti-IL-1β biologics. Additionally, neutrophil-NPs downregulate synoviocyte activation markers (CD248, fibronectin), thereby curbing synovial hyperplasia and invasiveness [84]. A series of macrophage membrane-based biomimetic nanoplatforms have been developed to enhance targeted therapy for RA by leveraging the natural inflammation-homing properties of macrophages and integrating multifunctional therapeutic mechanisms. Xu et al. [147] engineered a pathology-specific biomimetic multifunctional nanoplatform (M-M@I nanoparticles) combining mesoporous polydopamine (MPDA) for ROS scavenging, iguratimod (IGU) for anti-inflammatory and anti-osteoclast effects, and a macrophage membrane coating for targeted delivery. Through interactions between membrane-bound chemokine receptors (CCR2, CCR4) and adhesion molecules (CD44, Mac-1) with their counterparts (CCL2, CCL4, ICAM-1, P-selectin) at inflammatory sites, M-M@I accumulates in the paws of CIA mice, scavenges ROS, inhibits M1 macrophage polarization, reduces pro-inflammatory cytokines, suppresses bone erosion, and downregulates pathways such as NF-κB and IL-17, outperforming free IGU or non-coated formulations [147]. Xie et al. reported a biomimetic nanoplatform (MACP siTNF-α NPs) integrating Prussian blue nanoenzymes, TNF-α silencing siRNA, guanidinium-based polymers, and macrophage membranes. This system operates via a self-sustaining positive feedback loop in which PB scavenges ROS and restores glutathione, which degrades disulfide-containing polymers to release siRNA and expose more PB [148]. These nanoparticles are selectively taken up by M1 macrophages, reducing intracellular ROS and TNF-α expression, promoting polarization to the M2 phenotype, and, in CIA models, decreasing serum pro-inflammatory cytokines (TNF-α, IL-6, IL-1β), paw swelling, and bone erosion [148]. Hao et al. designed biomimetic nanoparticles coated with activated macrophage membranes (AMM) loaded with methotrexate (MTX), which target M1 macrophages via homotypic recognition [149], inducing apoptosis and polarization to M2, reducing TNF-α and IL-1β, and alleviating paw swelling and cartilage/bone destruction in adjuvant-induced arthritis (AIA) mice more effectively than platelet membrane (PM)-or naive macrophage membrane (MM)-coated nanoparticles [149]. Li et al. developed MMV-coated nanoparticles (MNPs) retaining CD44 and Mac-1, which interact with P-selectin and ICAM-1 on inflamed endothelium, enhancing accumulation in CIA joints [150], prolonging circulation, and effectively delivering tacrolimus to reduce inflammation, cytokine production (TNF-α, IL-1β), and bone erosion, outperforming bare or red blood cell membrane-coated nanoparticles [150]. Ma et al. highlighted other cytomembrane biomimetics: neutrophil membrane-coated nanoparticles inherit LFA-1, IL-1R, and TNF-αR, enabling binding to ICAM-1 on inflamed endothelium and adsorption of cytokines such as IL-1β and TNF-α, thereby reducing chondrocyte apoptosis, inhibiting MMP-3, and protecting cartilage [151]. Platelet membrane-coated vesicles utilize P-selectin and GPVI to target CD44 and collagen IV in inflamed synovium, improving delivery of immunosuppressants like FK506. This approach strategy achieves favorable blood retention and active targeting, thereby reducing systemic toxicity and improving therapeutic efficacy [151]. OX40 NVs, engineered to express OX40 receptors, demonstrate significant therapeutic effects in CIA models [146]. They accumulate in inflamed joints through OX40-OX40L interactions, reduce pathogenic CD4+OX40+ T cells, increase Tregs, and decrease pro-inflammatory cytokines such as TNF-α and IL-1β. In vivo, OX40 NVs alleviate joint swelling, lower arthritis scores, and mitigate cartilage destruction and bone erosion, exhibiting superior efficacy compared to anti-OX40L antibodies. Furthermore, dexamethasone-loaded OX40 NVs (DEX@OX40 NVs) produce synergistic effects by combining targeted drug delivery with OX40-OX40L blockade, further enhancing cartilage preservation and reducing bone loss [146]. Additional strategies have been explored to enhance delivery precision and therapeutic efficacy. N-PSAs target inflamed joints through LFA-1/ICAM-1 interactions, enabling sustained intra-articular release of salicylic acid, reducing NET formation, and protecting cartilage [113]. R4F-NM@F127 nanoparticles accumulate in RA joints via neutrophil-mediated chemotaxis and target macrophages through SR-B1-binding R4F peptides, reprogramming pro-inflammatory M1 macrophages to anti-inflammatory M2 phenotypes, decreasing TNF-α and IL-6 levels, and attenuating joint pathology [114]. NeuM-NPs combined with NSAIDs (delivered via microneedle patches) achieve superior joint targeting, cytokine suppression, and tissue preservation compared to uncoated nanoparticles [85]. Collectively, these studies underscore the versatility of neutrophil-derived vesicles as a targeted, multi-mechanistic treatment platform for RA with reduced systemic toxicity. Type 1 diabetes T1D is an autoimmune disease characterized by the selective destruction of pancreatic β-cells by autoreactive CD4+ and CD8+ T cells. A promising strategy to restore immune tolerance involves engineering RBC membrane-coated vesicles to display insulin peptides or β-cell-specific autoantigens. When administered to preclinical non-obese diabetic (NOD) mice, these vesicles induced antigen-specific tolerance through mechanisms such as clonal deletion, T cell anergy, and expansion of Tregs [80]. Prophylactic treatment delayed or prevented the onset of diabetes without the need for genetic engineering or systemic immunosuppression, offering a safe and scalable platform for tolerogenic immunotherapy. Furthermore, by leveraging the natural clearance pathway of RBCs via the reticuloendothelial system, allows these vesicles to mimic apoptotic cell processing and promote tolerogenic antigen presentation, thereby minimizing off-target immune activation [80]. Building on this approach, Wang et al. developed a bispecific cell-vesicle engager platform (aCD8-GrzBcs-IL2 MVs) designed to reverse newly diagnosed T1D [139]. These engineered vesicles co-display an anti-CD8 single-chain variable fragment (scFv) and a granzyme B-cleavable IL-2 mutant, enabling targeted suppression of autoreactive CD8+ T cells while locally activating Tregs. In NOD mouse models, intravenous administration reversed hyperglycemia in 77.8% of newly diagnosed cases without depleting peripheral CD3+ or CD8+ T cells, indicating a high safety profile. Controlled IL-2 release upon granzyme B cleavage ensures immune modulation is restricted to sites of β-cell attack, thereby avoiding systemic immunosuppression and enhancing treatment specificity [139]. In addition to systemic immune tolerance, local immune modulation is crucial for addressing complications associated with T1D, such as chronic wounds. A recent study developed macrophage membrane-coated nanoparticles loaded with the neddylation inhibitor MLN4924 (MM@MLN4924) to remodel the inflammatory microenvironment at wound sites [79]. These biomimetic vesicles accumulated in inflamed tissues, suppressed NF-κB activation, and promoted macrophage polarization toward the M2 phenotype. In diabetic mouse models, MM@MLN4924 accelerated wound closure and enhanced angiogenesis, demonstrating a dual therapeutic role in both immune regulation and tissue regeneration for T1D-related complications [79] (Fig. 5.). Schematic of Protein Neddylation Enzymatic Cascades and M-NPs/MLN4924 Preparation for Diabetic Wound Therapy (A) Schematic diagram illustrating the enzymatic cascade reactions involved in protein neddylation. The neddylation process proceeds through a sequence of key steps: first, NEDD8 is activated by the NEDD8 E1 activating enzyme (NAE); next, the activated NEDD8 is transferred to and loaded onto the NEDD8-specific E2 conjugating enzyme (UBC12); finally, with the assistance of NEDD8 E3 ligase, NEDD8 binds to specific protein substrates. Notably, the small molecule MLN4924 inhibits the neddylation process by suppressing the enzymatic activity of NAE. (B) Diagram showing the fabrication of M-NPs/MLN4924 and its therapeutic application in diabetic-induced wounds. In this study, a drug delivery system was constructed by coating NPs/MLN4924 with macrophage membranes, resulting in the formation of M-NPs/MLN4924. This nanoparticle system exerts multiple therapeutic effects: it regulates the proliferation and migration of endothelial cells, promotes angiogenesis, scavenges proinflammatory factors, and inhibits the polarization of macrophages toward the proinflammatory M1 phenotype, thereby facilitating diabetic wound healing. This figure was reproduced with permission from Ref. [79]. Copyright 2024, Elsevier. Autoimmune hemolytic anemia and immune thrombocytopenic purpura AIHA is a classic antibody-mediated disorder in which IgG autoantibodies target RBCs, leading to hemolysis. To mitigate this, RBC-ANS have been developed to act as decoys that adsorb circulating anti-RBC antibodies [55]. These nanosponges consist of PLGA nanoparticle cores cloaked with native RBC membranes, preserving surface antigenicity while enabling specific immunoadsorption. In murine models, RBC-ANS treatment significantly reduced hemolysis, restoring hematocrit and hemoglobin levels without the need for systemic immunosuppression [90]. This strategy offers high specificity and minimal off-target effects, providing a safer alternative to broad immunosuppressive therapies. Moreover, it may be adapted for other autoantibody-driven conditions by altering the membrane sources. A parallel strategy has been applied to ITP, an autoimmune disease characterized by antibody-mediated platelet destruction. Platelet membrane-coated nanosponges function as antibody decoys, sequestering anti-platelet autoantibodies while preserving hemostatic function [55]. These vesicles maintain platelet integrity by mimicking native antigenic profiles, thereby scavenging pathogenic antibodies without compromising physiological immunity. Wei et al. demonstrated that PNPs maintained approximately 70% of platelet counts and normalized bleeding times in murine models of ITP [92]. Both RBC-and platelet-based vesicle platforms are constructed using biodegradable PLGA cores, ensuring biosafety and biocompatibility for translational development. This biomimetic approach offers clear advantages over conventional therapies for AIHA and ITP-such as corticosteroids, intravenous immunoglobulin (IVIg), or splenectomy-by avoiding generalized immunosuppression while preserving immune specificity. Future directions may include developing hybrid membrane vesicles capable of scavenging multiple autoantibody species or co-delivering immunomodulatory agents to induce long-term tolerance. Collectively, these strategies collectively represent a next-generation therapeutic platform for antibody-mediated AIDs, with strong potential for clinical translation [55,92]. Multiple sclerosis MS is a chronic autoimmune disease characterized by demyelination and neuroinflammation within the central nervous system (CNS). Traditional therapies typically involve broad immunosuppression, which increases the risk of infections and compromises systemic immunity [152]. In contrast, cell-derived nanovesicles provide targeted, antigen-specific strategies that may overcome these limitations [153,154]. Initially developed for cancer immunotherapy, Cell-derived nanovesicles (CDNVs) have demonstrated the ability to present antigens to CD8+ T cells through cross-presentation by bystander DCs [49]. When engineered to carry myelin oligodendrocyte glycoprotein (MOG) or other CNS-specific antigens, CDNVs can potentially reprogram autoreactive T cells, inducing immune tolerance without broadly suppressing overall immune function. Furthermore, CDNVs can be functionalized with targeting ligands to direct them to lymphoid organs or inflamed CNS tissues, thereby enhancing therapeutic precision. However, future translational efforts must address challenges related to biodistribution, long-term stability, and safety in humanized models [49]. An innovative advancement in vesicle-based MS therapy is the development of mDCNVreg nanovesicles derived from epigenetically reprogrammed tolerogenic DCs [81] (Fig. 6A.). These vesicles exhibit high expression of MHC class II molecules while lacking costimulatory molecules such as CD80 and CD86, enabling direct antigen presentation to CD4+ T cells without the need for bystander DC assistance (Fig. 6B and C.). In EAE models, mDCNVreg loaded with the MOG35-55 peptide induced T cell anergy and expanded Tregs, effectively reducing neuroinflammation (Fig. 6D.). In CIA models, this system also preserves the host's defensive capacity against opportunistic infections (Fig. 6E.). Notably, these vesicles retain chemokines on their surface for specific lymphoid targeting and maintain phenotypic stability through IFN-γ mediated epigenetic programming. This platform circumvents systemic immunosuppression and overcomes limitations associated with adoptive cell therapy, offering a clinically viable approach for inducing durable antigen-specific tolerance in MS [81]. Schematic of Nanovesicle Engineering for Immunotherapy in Autoimmune Diseases (A) Schematic diagram illustrating the preparation process of membrane nanovesicles derived from DCs induced by IFN-γ. (B) Biodistribution pattern showing that the nanovesicles accumulate preferentially in draining lymph nodes. (C) Mechanism of action: The nanovesicles interact with T cells, promoting the differentiation of T cells into a tolerogenic phenotype. This process reestablishes antigen-specific immune tolerance and restores the balance of immune homeostasis. Therapeutic effects: Specifically, the induced Tregs migrate to inflammatory sites. This not only alleviates disease progression in both EAE. (D) and CIA. (E) models but also preserves the host's defensive capacity against opportunistic infections. This figure was reproduced with permission from Ref. [81]. Copyright 2025, ACS. Alter et al. reported a method to induce cell death, resulting in the production of large giant plasma membrane vesicles (GPMVs). These GPMVs are subsequently processed into monodisperse, smaller vesicles. This approach offers several advantages relevant to MS therapy [137]. The monodispersity of the resulting vesicles ensures a more consistent and reliable product for drug delivery and immunomodulation. In the context of MS, where effective homing and refined modulation of CNS immune responses are crucial, these monodisperse vesicles could be engineered to carry specific therapeutic payloads [137]. For example, they could potentially encapsulate immunomodulatory drugs or agents that can promote remyelination. Given the challenges of delivering drugs across the blood-brain barrier (BBB) in MS treatment, the unique properties of these EVs might be exploited. Their natural origin from the plasma membrane may confer enhanced BBB-crossing capabilities [137]. Yang et al. developed an inflammation-targeted biomimetic nano-decoy (TFMN) to enhance MS treatment [104]. TFMN is constructed by coating neutrophil membranes (NM) onto nanoparticles self-assembled from tannic acid and poloxamer188, encapsulating methylprednisolone (MPS). Leveraging the natural inflammation tropism of activated NM, TFMN targets CNS lesions, adsorbs neutrophil-related cytokines to reduce infiltration, and modulates the inflammatory microenvironment in EAE mice [104]. It scavenges excessive ROS to provide neuroprotection. At inflammation sites, perforin released from cytotoxic T lymphocytes triggers controlled MPS release through NM pores, optimizing drug accumulation while protecting neurons from perforin toxicity. TFMN regulates T lymphocyte subsets by reducing pathogenic Th1/Th17 cells and increasing Tregs, while lowering pro-inflammatory cytokines such as IFN-γ and IL-17, offering a novel strategy for MS treatment [104]. Autoimmune myocarditis Autoimmune myocarditis is characterized by immune-mediated cardiomyocyte injury, diffuse myocardial inflammation, and progressive cardiac dysfunction, which can ultimately progress to dilated cardiomyopathy and heart failure [155]. Current systemic immunosuppressive regimens often cause broad off-target toxicity and fail to achieve sufficient local drug concentration in cardiac tissue, highlighting the critical need for targeted, minimally invasive therapeutic delivery strategies for cardiac inflammation [156]. IACMVs have emerged as promising therapeutic tools for autoimmune myocarditis due to their ability to target inflamed tissues, modulate pathogenic immune responses, and incorporate diagnostic functionalities [99,100,157]. IACMV-based nanoparticles serve as multifunctional platforms for imaging and treatment of autoimmune myocarditis. Macrophage membrane-modified nanoparticles (MM/NPs) are fabricated by coating PLGA cores, loaded with perfluorohexane (PFH) and superparamagnetic iron oxide (SPIO), with macrophage membranes expressing integrin α4β1 and CD47 [157]. Integrin α4β1 mediates active targeting to inflamed myocardial tissue via binding local inflammatory chemokines, while CD47 confers immune evasion [157]. In experimental autoimmune myocarditis (EAM) models, MM/NPs accumulate selectively in cardiac lesions to support ultrasound, photoacoustic and magnetic resonance multimodal imaging for early diagnosis, and suppress pro-inflammatory macrophage activation to reduce TNF-α and IL-6 release, thereby resolving myocardial inflammation and preserving cardiac function [157]. For enhanced targeting and immunomodulatory performance, macrophage-T cell hybrid membrane nanoparticles (siIRF1@ZIF@HM NPs) were engineered to carry both macrophage-derived ITGA4/ITGB1 and T cell-derived TCR-β, which specifically target pro-inflammatory M1 macrophages through the CXCL9/10/11-CXCR3 axis [99,100]. Within a pH-responsive zeolitic imidazolate framework-8 (ZIF-8) core encapsulating IRF1-siRNA, these nanoparticles realize efficient endosomal escape and gene silencing; IRF1 knockdown inhibits macrophage pyroptosis, downregulates GSDMD-NT and cleaved Caspase-1, and reduces secretion of IL-1β and TNF-α [99,100]. In EAM models, siIRF1@ZIF@HM NPs mitigate myocardial inflammationand leukocyte infiltration, decrease serum CK-MB levels, and protect cardiac function without causing systemic toxicity [99,100]. Despite the favorable inflammation-homing properties of these IACMV-based nanotherapeutics, systemic intravenous delivery often yields low myocardial accumulation due to cardiac barrier limitations and rapid systemic clearance [158,159]. To address this translational bottleneck, localized delivery strategies, including intrapericardial injection and transcutaneous cardiac microneedle patches, have been adopted to achieve direct, long-term, site-specific IACMV delivery, thereby improving cardiac bioavailability, minimizing systemic off-target immunosuppression, and promoting the clinical translation of vesicle-based cardiac therapy [[160], [161], [162], [163]]. Consistent with this advancing trend of localized and minimally invasive cardiac treatment, a series of emerging biomaterials and next-generation vesicle engineering platforms have further broadened the toolkit for cardiac immune modulation and tissue repair [[163], [164], [165], [166], [167]]. Representative strategies include inhalable stem cell-derived exosomes, immunomodulatory microneedle systems, and stromal platelet membrane-inspired nanoparticles, which achieve efficient cardiac inflammation resolution and tissue repair through distinct delivery and targeting designs [164,165]. Plant-derived functional hydrogels and optimized cardiac delivery methodologies also provide valuable translational guidance for the rational design of cardiac therapeutics [166,167]. Together, these localized delivery approaches complement the inherent targeting advantages of immune cell-derived vesicles, improving cardiac bioavailability, reducing off-target immunosuppression, and advancing the clinical translatability of IACMV-based therapy for autoimmune myocarditis. Collectively, single-source and hybrid IACMV platforms demonstrate unique advantages in targeted inflammation resolution and immune regulation, providing integrated diagnostic and therapeutic strategies for the complex immune microenvironment of autoimmune myocarditis. Combined with emerging localized delivery technologies and advanced biomaterial-based cardiac therapeutic systems, IACMV-based nanotherapeutics hold great promise for future clinical translation in treating refractory cardiac inflammatory diseases. In Table 5, we have summarized the therapeutic applications and efficacy of IACMVs in various AIDs. Table 5Therapeutic applications and efficacy of IACMVs in AIDs.Table 5Autoimmune Disease TypeIACMVs -Based TherapiesKey Findings/Mechanisms of ActionReferencesIBDMacrophage-derived vesicles (e.g., Motor@M2M@SAM, Ber@MVs-CA)Targeted accumulation in inflamed intestine; scavenged TNF-α, IL-6; promoted M2 polarization; restored epithelial barrier (ZO-1, occludin); pH-responsive drug release[58,59,78]NM-PBCombined neutrophil targeting (LFA-1/ICAM-1) with macrophage immunoregulation; scavenged ROS; scavenged TNF-α, IL-6; promoted M2 polarization[97]OX40 NVsAccumulated in inflamed colon via OX40-OX40L interaction; reduced pathogenic CD4+OX40+ T cells; increased Tregs; decreased TNF-α, IL-1β; alleviated weight loss, improved DAI, restored colon length[146]MSN-HC@NM NPspH-responsive drug release (∼65% at pH 5.0); reduced TNF-α, IL-1β; restored tight junction proteins; superior histopathological improvement in DSS-colitis[86]PNPsTargeted damaged colonic vasculature; reduced neutrophil infiltration, IL-1β, TNF-α; restored epithelial integrity; inhibited Ca2+-ROS signaling[94]KGF-NeusEnhanced KGF stability; targeted inflamed sites; reduced intestinal mucosal ulcers and neutrophil infiltration; promoted recovery of intestinal morphology[144]LM-CsA NPs/MM-CsA NPsUtilized integrins α4β1/α4β7 to target VCAM-1/MadCAM-1; prolonged retention in inflamed intestines; mitigated weight loss, colon shortening, DAIs; decreased TNF-α, IL-6, MPO; increased E-cadherin, ZO-1, occludin.[145]SLERe-BCMA-NVsScavenged APRIL and BAFF; suppressed NF-κB signaling[133]CD40 NVs, MMFReduced anti-dsDNA and ANA autoantibodies; improved renal function; limited immune complex deposition[45]MPA-R-LipoSuppressed DC maturation (downregulated CD40, CD80); reduced TNF-α, IL-12p70; alleviated lupus nephritis in MRL/lpr mice[87]CsA-RNPsImproved pharmacokinetic profiles (2.05-fold longer half-life); decreased levels of anti-dsDNA antibodies and inflammatory cytokines; alleviated lupus nephritis.[88]Neutrophil-derived PMVsEnriched with phosphatidylserine; mimicked apoptotic bodies; induced tolerance via silent phagocytosis[137]IM-MNPs/DXMDual targeting of PD-1 and TIGIT on pathogenic CD4+ T cells; suppressed effector T cells; enhanced Treg function; alleviated lupus nephritis; targeted organ delivery[136]Decoy vesiclesCytosolic DNA in AdMVs activates cGAS-STING-IFN-I; decoy vesicles lacking DNA could competitively inhibit cGAS.[52]Reprogrammed apoptotic vesiclesPotential to act as decoys to scavenge autoantibodies or cytokines, redirecting immune responses toward tolerance[54]RAFK506-loaded PNPsTargeted inflamed synovium via P-selectin, GPVI interactions with CD44, collagen IV; enhanced joint accumulation; reduced joint inflammation, bone erosion, cytokine production (TNF-α, IL-6, IL-1β)[93]Neutrophil-NPsRetained LFA-1, IL-1R, TNF-αR; targeted inflamed synovium; sequestered IL-1β, TNF-α; suppressed chondrocyte apoptosis; inhibited MMP-3; reduced joint swelling, leukocyte infiltration, cartilage degradation[84]DEX@OX40 NVsAccumulated in inflamed joints via OX40-OX40L; reduced pathogenic CD4+OX40+ T cells; increased Tregs; decreased TNF-α, IL-1β; alleviated joint swelling, cartilage destruction, bone erosion; synergistic with dexamethasone[146]N-PSAsTargeted joints via LFA-1/ICAM-1; sustained intra-articular salicylic acid release; inhibited NET formation (via MAPK pathway); promoted cartilage repair[113]R4F-NM@F127Accumulated in joints via neutrophil chemotaxis (IL-8) and LFA-1; targeted SR-B1 on synovial macrophages; reprogrammed M1 to M2; decreased TNF-α, IL-6[114]NeuM-NPsPreserved TNF-αR, IL-1βR; high cytokine binding affinity; enhanced accumulation in inflamed paws; superior joint targeting, cytokine suppression, tissue preservation[85]Leukocyte membrane-coated nanoparticlesPreferentially accumulated in inflammation sites; reduced TNF-α, IL-6; mitigated tissue damage[52]Macrophage membrane-based nanoplatforms (e.g., M-M@I, MACP siTNF-α NPs, AMM-MTX NPs, MMVs-coated MNPs)Leveraged macrophage homing (CCR2/CCR4, CD44/Mac-1); scavenged ROS; promoted M2 polarization; suppressed cytokines (TNF-α, IL-6, IL-1β); reduced bone erosion; improved targeting and efficacy[[147], [148], [149], [150]]T1DErythrocyte membrane-coated vesiclesInduced antigen-specific tolerance (clonal deletion, T cell anergy, Treg expansion) in NOD mice; delayed/prevented diabetes onset; mimicked apoptotic cell clearance[80]aCD8-GrzBcs-IL2 MVsCo-displayed anti-CD8 scFv and granzyme B-cleavable IL-2 mutant; targeted suppression of autoreactive CD8+ T cells; local Treg activation; reversed hyperglycemia in 77.8% of newly diagnosed NOD mice[139]MM@MLN4924Accumulated in inflamed tissues; suppressed NF-κB activation; promoted M2 polarization; accelerated diabetic wound closure; enhanced angiogenesis[79]AIHARBC-ANSPLGA cores cloaked with native RBC membranes; preserved antigenicity; adsorbed circulating anti-RBC IgG; prevented hemolysis; restored hematocrit and hemoglobin without systemic immunosuppression[90]ITPPNPsServed as decoys for anti-platelet antibodies; preserved platelet surface antigens; scavenged pathogenic IgG; maintained ∼70% platelet counts; normalized bleeding time in murine ITP[92]MSmDCNVregPotential for antigen presentation and cross-presentation by bystander DCs; reprogram autoreactive T cells; induce immune tolerance[49]mDCNVregHigh MHC class II, low CD80/CD86; induced CD4+ T cell anergy and Treg expansion; alleviated neuroinflammation in EAE; preferential lymphoid targeting; phenotypic stability via IFN-γ programming[81]nPMVs from GPMVsPotential for enhanced BBB crossing and CNS drug delivery[137]TFMNTargeted CNS lesions via NM tropism; adsorbed neutrophil-related cytokines; scavenged ROS; perforin-triggered MPS release; reduced Th1/Th17, increased Tregs; lowered IFN-γ, IL-17[104]AMMM/NPs loaded with PFH/SPIOTargeted inflamed myocardium via integrin α4β1; immune evasion via CD47; enabled multimodal imaging (US/PA/MRI); suppressed TNF-α, IL-6; promoted inflammation resolution[157]siIRF1@ZIF@HM NPsTargeted M1 macrophages via CXCL9/10/11-CXCR3 axis (ITGA4/ITGB1, TCR-β); pH-responsive ZIF-8 core for endosomal escape and IRF1-siRNA delivery; suppressed macrophage pyroptosis (reduced GSDMD-NT, cleaved Caspase-1, IL-1β, TNF-α); alleviated myocardial inflammation, preserved cardiac function in EAM[99,100] Therapeutic applications and efficacy of IACMVs in AIDs. Challenges and future perspectives of IACMVs in autoimmune disease therapy IACMVs have emerged as promising biomimetic platforms for targeted and immune-modulatory therapy of autoimmune diseases, owing to their inherent cell-mimicking properties, favorable biocompatibility, and tunable immune regulatory functions [168]. Nevertheless, substantial translational challenges including heterogeneous immunogenicity, unstable large-scale manufacturability, variable targeting capacity, and insufficient long-term safety evaluation still limit their clinical transformation [169]. The subtype-specific differences in immunogenic characteristics, production scalability, and targeting performance of typical IACMVs are systematically summarized in Table 6. Table 6Comparison of immunogenicity, scalability and core traits of diverse IACMVs.Table 6Vesicle subtypeCore immunogenic characteristicsImmunogenicity levelScalability & manufacturabilityReferencesMMVsExhibit polarization-dependent immunogenicity: pro-inflammatory M1-type MVs carry abundant DAMPs to trigger innate immune activation, while anti-inflammatory M2-type MVs possess immune-regulatory properties with low pro-inflammatory riskHeterogeneous (M1: High; M2: Low)Moderate yield; strongly affected by M1/M2 polarization state, causing significant batch-to-batch inconsistency[[170], [171], [172]]NMVsSeveral MVs populations carry pro-inflammatory miRNAs and proteases that strongly activate endothelial cells and monocytes (e.g., miR-142-3p, miR-155, MMP-8/9), but other subsets (e.g., NDMVs, LAND-Vs) can exert anti-inflammatory and complement-regulatory effects.Heterogeneous(Pro-inflammatory High; some anti-inflammatory subtypes)High yield via nitrogen cavitation/membrane-engineering methods; scalable preparation demonstrated, but long-term storage stability and industrial-scale robustness still require systematic evaluation[[173], [174], [175]]DCMVsCarry functional MHC-I/II–peptide complexes and co-stimulatory molecules (e.g., CD80, CD86), enabling potent activation of antigen-specific CD4+ and CD8+ T cells and NK cells; can be loaded with tumor or pathogen antigens to induce strong adaptive immune responses and cross-presentation.Moderate–High (strong antigen-specific T-cell priming and Th1-biased responses when derived from mature/adjuvant-treated DCs).Natural DC MVs have limited yield and complex production; DC membrane-derived nanovesicles generated by nitrogen cavitation can be produced in large yields, overcoming major production constraints for vesicle-based vaccines.[[176], [177], [178]]PLT-MVsAbundantly released by activated or apoptotic platelets and loaded with coagulation factors, cytokines, and ROS. They exert potent procoagulant effects and context-dependent pro-/anti-inflammatory and immunomodulatory activities.Moderate-High (marked procoagulant and thrombo-inflammatory potential, with additional immunoregulatory and reparative functions depending on stimulus and cargo).Highly abundant in circulation and easily prepared via platelet activation or lysis, supporting scalable production; unified operational protocols remain to be standardized.[123,[179], [180], [181]]RBC-MVsLack classical antigen presentation and overexpress CD47 and other “self” markers, conferring intrinsic immune inertness and efficient evasion of macrophage phagocytosis.Low (immune-evasive, with minimal risk of adaptive immune activation under physiological conditions).High-yield from abundant RBC sources, but true large-scale standardized production remains technically challenging.[[182], [183], [184]]Hybrid VesiclesIntegrate multiple membrane sources (e.g., cellular, bacterial, plant or synthetic) to co-display antigens, adjuvants and self-markers, enabling rational tuning of immunogenicity from low-immunostimulatory carriers to highly immunogenic vaccine platforms via membrane composition and fusion ratio control.Tunable (from low to high, depending on parental membranes, cargo, and fusion design).Process-dependent scalability; membrane fusion method and composition (e.g., extrusion vs sonication, membrane/protein ratios) critically influence batch-to-batch reproducibility, physical stability, and production yield.[[185], [186], [187]] Comparison of immunogenicity, scalability and core traits of diverse IACMVs. Cell source-dependent immunogenicity heterogeneity Immunogenicity heterogeneity is the foremost challenge restricting the clinical application of IACMVs in autoimmune therapy [188]. Unlike synthetic nanomaterials with uniform properties, the immune reactivity of IACMVs is highly dependent on their parental cell types, cell activation status, and polarization phenotypes, resulting in distinct immune risks among different vesicle subtypes (Table 6). In general, macrophage and neutrophil-derived MVs exhibit highly heterogeneous immunogenicity. Their pro-inflammatory subsets may trigger innate immune activation and inflammatory exacerbation via DAMPs and pro-inflammatory nucleic acids, whereas specific anti-inflammatory subpopulations exert immune-regulatory effects [[189], [190], [191]]. DC-derived MVs possess moderate-to-high immunogenicity due to retained MHC-peptide complexes and co-stimulatory molecules, which may initiate unwanted adaptive immune priming [112]. Platelet-derived MVs display moderate-to-high thrombo-inflammatory potential with context-dependent immunomodulatory functions [192]. In contrast, RBC-derived MVs maintain low immunogenicity owing to intrinsic immune-evasive characteristics [193]. Notably, hybrid MVs enable fully tunable immunogenicity through rational membrane combination and component optimization, providing great potential for refined immune modulation [194]. Future studies should focus on optimizing vesicle surface composition and nucleic acid cargo purification to develop immuno-silent and immuno-tunable IACMVs for safe autoimmune intervention. Beyond the well-recognized cGAS-STING-mediated interferon response triggered by vesicle-associated nucleic acids [53], repeated long-term administration of IACMVs may induce the production of anti-drug antibodies (ADAs) against exposed membrane proteins or engineered surface moieties [168,188]. Such adaptive humoral responses can scavenge vesicle therapeutic efficacy, accelerate in-vivo clearance, and even trigger hypersensitivity reactions, particularly critical for chronic autoimmune disorders requiring continuous intervention [195]. Future studies should focus on optimizing vesicle surface composition, conducting nucleic acid cargo purification, and developing stealth-coated or self-marker-modified immuno-silent IACMVs to minimize both innate inflammatory activation and ADA formation for safe long-term autoimmune intervention. Scalability and batch-to-batch reproducibility Manufacturing scalability and batch stability constitute critical technical bottlenecks for the industrial translation of IACMVs. As illustrated in Table 6, different IACMV subtypes possess vastly distinct production feasibility and reproducibility. Neutrophil and platelet-derived MVs support relatively high-yield preparation, while RBC-derived MVs benefit from abundant cell sources. However, the production stability of these vesicles is still restricted by non-standardized preparation protocols and storage limitations [[196], [197], [198]]. Macrophage-derived MVs suffer from prominent batch variability caused by unstable M1/M2 polarization states. Natural DC-derived MVs are severely limited by low yield and complex preparation procedures, though advanced membrane extrusion strategies have partially overcome such restrictions [176,199]. For hybrid MVs, production yield and quality stability are highly process-dependent, where fusion methods and membrane ratios profoundly affect batch consistency [200]. To address low vesicle yield, upstream cell engineering and high-secretion subpopulation sorting have emerged as actionable pre-production optimization strategies. As demonstrated by Koo et al., selective isolation of parental cell subpopulations with inherently high extracellular vesicle secretion capacity can substantially boost per-cell vesicle output, while targeted genetic or epigenetic engineering of source cells further enhances membrane vesicle release efficiency, directly improving overall manufacturing scalability [201]. Notably, donor-to-donor variability represents a largely under-recognized confounding factor that disproportionately impacts MMVs and DCMVs. Macrophage polarization profiles and DC antigen-presentation capacity are highly sensitive to donor genetics, age, health status, and ex-vivo culture conditions, leading to inconsistent functional phenotypes across donor batches and limiting therapeutic reproducibility [202,203]. Current quality control for IACMVs mainly relies on basic physical indicators including particle size distribution and zeta potential, which are insufficient to guarantee functional consistency [204]. Critical quality attributes (CQAs) closely associated with therapeutic performance, such as copy numbers of key functional membrane proteins (e.g., adhesion receptors, immune checkpoint molecules), surface marker density, and cargo-loading efficiency as well as in-vitro drug release profiles, are often neglected during batch evaluation [205,206]. Accordingly, future Good Manufacturing Practice (GMP)-relevant release criteria for clinical-grade IACMVs should integrate not only physicochemical parameters but also functional readouts, including polarization status for MMVs, antigen-presentation activity for DCMVs, and standardized potency assays to minimize donor-associated heterogeneity. Collectively, current IACMV preparation largely relies on laboratory manual operation. Establishing GMP-compliant standardized workflows, automated bioprocessing platforms, and unified quality control criteria is essential to achieve scalable and reproducible clinical-grade production. Targeting specificity and rational engineering strategies Targeting specificity determines the therapeutic accuracy and off-target risk of IACMV-based autoimmune therapy. Each vesicle subtype inherits unique biological tropism from its parental cells, showing differential lesion-targeting capabilities. Neutrophil and macrophage MVs exhibit excellent inflammation tropism toward acute inflammatory and repairing tissues [207,208], while platelet MVs efficiently target injured vasculature and thrombo-inflammatory lesions [209]. DC-derived MVs possess unique antigen-specific immune recognition ability, enabling specific adaptive immune regulation [110]. In contrast, RBC-derived MVs mainly rely on passive long-circulation effects with limited active targeting capacity [210]. Despite inherent targeting advantages, most native IACMVs exhibit limited preferential accumulation at disease lesions. Notably, current relevant studies generally lack quantitative biodistribution data to systematically assess off-target accumulation in organs such as the liver and spleen, and a considerable portion of administered dose tends to distribute to these non-target tissues, which is an unaddressed limitation for further translational research [211,212]. Advanced rational engineering strategies can further improve their targeting performance. Surface modification with disease-specific ligands and immune checkpoint molecules effectively enhances lesion selectivity. Hybrid membrane design integrates multiple cellular tropisms to achieve dual-targeting advantages [213]. Artificial intelligence (AI) and machine learning have emerged as powerful tools to move beyond empirical trial-and-error modification toward data-driven rational design [214]. For instance, structure-based machine learning models can predict optimal spatial orientations of membrane-embedded receptors and ligands to maximize target binding efficiency while avoiding steric hindrance [215]. Quantitative structure-activity relationship (QSAR)-based algorithms can further calculate ligand-receptor binding affinities and screen optimal surface modification combinations tailored to inflammatory microenvironments [216,217]. Long-term safety, pharmacokinetics and regulatory challenges Despite the favorable biocompatibility of IACMVs observed in short-term preclinical studies, systematic evaluation of long-term in vivo safety, pharmacokinetic behavior, and repeated administration toxicity remains insufficient. Residual chemical impurities introduced during vesicle functionalization and modification may cause potential biosafety risks [137]. Advanced engineering technologies such as CRISPR editing and modular bioconjugation further enrich vesicle functions but also bring uncertain long-term safety concerns [218]. Machine learning algorithms can assist pharmacokinetic modeling and biodistribution prediction by analyzing vesicle physical properties and surface modification parameters, guiding the optimization of circulation half-life and in vivo accumulation profiles [219,220]. In addition, newly emerging hybrid and synthetic biomimetic vesicles lack standardized regulatory guidelines, which hinders their clinical transformation. In the future, comprehensive long-term in vivo validation focusing on ADA-related adverse effects, standardized manufacturing system construction, and early regulatory communication are urgently required to promote the clinical application of next-generation IACMV therapeutics for autoimmune diseases [221,222]. Future perspectives Against the backdrop of rapid development of extracellular vesicle engineering and biomimetic nanotechnology, continuous technical innovation will further expand the application boundary of IACMVs in autoimmune disease treatment. Drawing lessons from cutting-edge progress in the broader vesicle field will become an important direction to break through existing technical bottlenecks. Engineered exosome-like nanovesicles, which are deliberately designed to strengthen tissue targeting and prolong in vivo retention, provide valuable references for the functional upgrading of IACMVs [223]. Inspired by decoy exosome systems that efficiently sequester inflammatory mediators and novel strategies targeting pathological extracellular vesicles, researchers can further expand the working modes of IACMVs to intervene in complex inflammatory and immune networks [224,225]. Beyond functional modification and mechanism expansion, upstream optimization of cellular sources and high-precision characterization technologies will also strongly support the standardized development of IACMV therapeutics. Rational screening and optimization of vesicle-producing cell lines help stabilize vesicle properties and improve overall production performance [201]. Meanwhile, emerging single-vesicle analysis techniques enable in-depth evaluation of vesicle morphology, membrane composition and biological activity at the individual particle level, which complements conventional bulk detection methods and facilitates the establishment of more rigorous quality control systems [226]. In the long run, integrating advanced vesicle engineering, high-resolution characterization, and multi-omics analysis will drive IACMVs to evolve from simple biomimetic carriers into intelligent, multi-functional and precisely controllable therapeutic platforms. Such interdisciplinary innovations will lay a solid foundation for translating immune cell membrane vesicles into reliable clinical therapies for refractory autoimmune diseases. Conclusion IACMVs present a promising approach for treating autoimmune diseases by integrating biocompatibility, immune regulation, and targeted delivery within a single platform. Preclinical studies have demonstrated their capacity to scavenge inflammatory mediators, restore immune tolerance, and deliver therapeutic agents with high precision. Despite these advances, challenges such as immunogenicity, manufacturing scalability, and long-term safety remain significant obstacles to clinical translation. Future efforts should focus on optimizing vesicle design, enhancing targeting specificity, and establishing standardized production protocols. With continued interdisciplinary innovation, IACMVs are poised to become next-generation tools for personalized and tolerogenic therapies in autoimmune disease management. Data availability statement Not applicable. Ethics approval and consent to participate Not applicable. This study did not involve human subjects or animal experiments. CRediT authorship contribution statement Yan Li: Conceptualization, Data curation, Writing – original draft. Nasrullah Jan: Conceptualization, Data curation, Software. Jimin Zhang: Formal analysis, Investigation. Yang Qin: Resources, Software. Gang Liu: Supervision, Writing – review & editing. Chao Liu: Supervision, Writing – review & editing. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work was sponsored by the National Natural Science Foundation of China (82272144), the Guangdong Basic and Applied Basic Research Foundation (2025B1515020074), and the Shenzhen Science and Technology Innovation Commission (JCYJ20220530143411025, JCYJ20230807091404010, JCYJ20250526161328004), the Natural Science Foundation of Fujian Province (2025J09009), and the Scientific Research Foundation of State Key Laboratory of Vaccines for Infectious Diseases, Xiang An Biomedicine Laboratory (2024XAKJ0102004). Some scientific images were generated using ScienceSlides (Visiscience.com, Raleigh, NC, USA). Footnotes Peer review under the responsibility of editorial board of Bioactive Materials. Contributor Information Gang Liu, Email: gangliu.cmitm@xmu.edu.cn. Chao Liu, Email: liuchao66888@xmu.edu.cn. References Associated Data Data Availability Statement Not applicable.