INTRODUCTION

Autoimmune diseases are chronic, debilitating, incurable, and life-threatening diseases; patients with autoimmune diseases need to receive treatments throughout their life. Patients with systemic lupus erythematosus (SLE) may have inflammation and tissue damage in the liver, kidney, skin, lung, joint, central nervous system, and other organs (1). Despite recent advances in biologic therapies (such as tocilizumab/anti–interleukin-6 receptor [anti–IL-6R] antibody and adalimumab/anti–tumor necrosis factor [anti-TNF] antibody), 30% of patients with rheumatoid arthritis (RA) and 26–38% of patients with ankylosing spondylitis fail to respond to all therapies (2-4). Furthermore, the majority of patients who show improvement after treatment do not achieve complete remission, and their response to therapy may diminish over time (5, 6). Diagnosis and treatment of SLE are challenging due to complex symptoms and a lack of effective therapeutics (1, 6). Identification of novel therapeutic targets will help future development of effective treatments for SLE. Moreover, novel diagnostic/prognostic biomarkers will help to stratify patients who are likely to respond to a specific drug, leading to precision medicine.

T cells promote autoimmune diseases by inducing autoantibody production and inflammatory responses (7-12). Effector memory T cell and Th17 cell numbers are increased in SLE patients (7, 13, 14). The Th1:Th2 cell ratio is also enhanced in SLE patients (15, 16). Th1-secreted interferon-γ (IFNγ) and TNF contribute to macrophage activation and damage of multiple tissue types (15). Th17-secreted IL-17A is a key pathogenic cytokine in inflammation and autoimmune responses (17, 18). Th17 cells recruit macrophages and dendritic cells to inflammation sites; Th17 cells also facilitate B cell activation and autoantibody production (17). Conversely, the Treg cell population is decreased in SLE patients (19). Thus, T cell hyperactivation plays a critical role in the pathogenesis of SLE.

Cell-derived exosomes directionally deliver proteins, amino acids, microRNA, or metabolites to targeted cells or tissues to modulate cell or tissue characteristics (20-23). Moreover, T cell–derived exosomal microRNAs modulate immune responses (22, 24, 25). The number of exosomes in the sera of SLE patients is correlated with disease severity (26). These serum-derived exosomes from SLE patients induce the production of proinflammatory cytokines by impacted peripheral blood mononuclear cells from healthy individuals (26). To date, the surface proteins and intra-exosomal proteins of exosomes in SLE patients, as well as the regulatory mechanisms of exosomal protein–induced inflammation in SLE patients, remain unclear.

Eosinophil cationic protein (ECP; also called human RNase III) is a defense protein; eosinophils release ECP during degranulation against bacterial or parasitic infection (27). ECP disrupts the bacteria membrane through binding to lipopolysaccharide or other bacterial cell wall components (27). In addition to being increased during infection, ECP levels are increased in human patients with allergic asthma or atopic dermatitis (28). Moreover, ECP treatment induces mammalian cell necrosis and inhibits cell growth and proliferation (29-31). ECP treatment also induces cell apoptosis through TNF–caspase signaling (32). To date, the roles of ECP in T cell function and autoimmune disease pathogenesis remain unknown. In this study, we characterized T cell–derived exosomes from SLE patients and identified ECP as a pathogenic exosomal protein.

DISCUSSION

A key finding of this study was the identification of one novel T cell exosomal protein, ECP, which plays an important role in SLE pathogenesis. ECP overexpression in T cells resulted in enhancement of inflammatory responses and T cell activation. Notably, ECP-containing exosomes from T cells targeted several tissues (such as the liver, kidney, and joint) of the recipient mice, leading to tissue inflammation. These data suggest that ECP-overexpressing T cells or ECP-containing exosomes may act as a causal factor in SLE.

One of the notable findings of this study is that T cell–derived exosomal ECP contributes to autoimmune diseases. This is the first study to show that ECP is an exosomal protein. Extracellular ECP stimulation using ECP recombinant protein induces cell death, including necrosis and apoptosis (29, 31, 32); however, T cell development in Lck-ECP–transgenic mice was not affected. The data suggest that ECP recombinant protein acts differently from exosomal and intracellular ECP. In support of this notion, soluble ECP levels were not increased in the sera of human SLE patients. These findings also suggest that exosomal ECP and ECP-overexpressing T cells contribute to autoimmune responses through cell death–independent pathways. Furthermore, adoptive transfer of ECP-containing exosomes induced autoantibody production and inflammation expansion; it is likely that the T cell–derived exosomes from Lck-ECP–transgenic mouse T cells contain other inflammatory molecules in addition to ECP proteins.

Besides induction of the proinflammatory cytokine IFNγ, Lck-ECP–transgenic mice manifested the induction of Tfh cells, plasma B cells, and autoantibodies. These results suggest that ECP overexpression in T cells induces the Tfh cell population, facilitating plasma B cell differentiation, leading to overproduction of autoantibodies. Besides induction of plasma B cells through Tfh cells, it is also possible that secreted molecules from ECP-overexpressing T cells or other proteins within ECP+ exosomes may stimulate/activate other potential target cells (e.g., B cells, macrophages, dendritic cells, or osteoclasts), leading to multiple inflammatory phenotypes.

The aforementioned findings are consistent with our single-cell RNA-sequencing data using Lck-ECP–transgenic mouse T cells. First, Lck-ECP–transgenic mouse T cells showed highly increased levels of GH, which induces T cell survival and activation (37, 38). Consistent with these findings, T cell activation, T cell proliferation, and adaptive immune response pathways were indeed induced in Lck-ECP–transgenic mouse T cells. Lck-ECP–transgenic mouse T cells also showed induction of IFNγ+ Th1 differentiation.

Second, ECP signaling induced several proinflammatory cytokines/chemokines, including TNFSF8 (also called CD30 ligand [CD30L]), S100 proteins, and IFNγ, that may contribute to inflammation of multiple tissue types. CD30L up-regulation is involved in the pathogenesis of human SLE, RA, Hodgkin lymphoma, and anaplastic large cell lymphoma (39). S100 protein enhancement contributes to arthritis and neural degenerative diseases (40, 41). Chronic IFNγ overproduction induces hepatoxicity (42, 43); IFNγ also plays a crucial role in the development of nephritis (44, 45). Interestingly, our ELISA data also showed early induction of IFNγ in Lck-ECP–transgenic mice. These previous publications and our data suggest that ECP-overexpressing T cells and ECP-containing exosomes cooperate with the aforementioned proinflammatory cytokines/chemokines to induce nephritis, arthritis, and hepatitis in Lck-ECP–transgenic mice and maybe also in human SLE patients.

Third, surface receptors on exosomes may determine their tissue tropisms. It would be interesting to study whether the exosomal olfactory receptor 7D2, identified in this study as being enriched in SLE (Supplementary Table 2), controls the tissue tropism of inflammatory T cell exosomes. Finally, the role of ECP-inducible Titin, an intrasarcomeric filamentous protein, in SLE pathogenesis needs to be explored.

Taken together, these findings indicate that ECP overexpression in T cells contributes to inflammatory responses through both intrinsic and extrinsic events. To distinguish the specific role of ECP in T cells versus exosomes is highly challenging, if not impossible, due to the following two reasons. First, there is no definitive ECP orthologous gene (46) for the generation of ECP-knockout mice. Second, there are no exosome-specific surface markers for the depletion of exosomes in vivo.

SLE patients experience damage to multiple organs and complex symptoms (1). Understanding the causal factors of individual symptoms will help in the development of novel therapeutics for SLE. In this study, we found that ECP-containing exosomes induce arthritis, hepatitis, and nephritis in mice. According to our clinical data and proteomics analysis, 5 SLE patients had ECP-containing exosomes derived from T cells, while all of these 5 SLE patients developed arthritis. The data suggest that the novel pathogenic factor, ECP-containing exosome, may also be a biomarker for SLE-associated arthritis. In addition, 2 of these 5 SLE patients have developed nephritis; however, it is difficult to diagnosis SLE-associated nephritis at an early stage. ECP-containing exosomes may help in the early diagnosis of SLE-associated nephritis. Although up to 50% of SLE patients develop hepatitis, the exact diagnosis of hepatitis remains challenging due to complex conditions, including infection, drug treatment, or SLE (47). Notably, Lck-ECP–transgenic mice spontaneously developed hepatitis, suggesting that hepatitis in SLE patients may be a consequence of the induction of ECP-containing exosomes. Thus, exosomal ECP may also be a potential biomarker for SLE-associated hepatitis.

Taken together, our findings suggest that ECP overexpression in T cells or T cell–derived exosomes induces T cell hyperactivation and proinflammatory cytokine production through both intrinsic and extrinsic events, leading to inflammation in multiple organs and autoimmune responses. Thus, ECP-overexpressing T cells and ECP-containing exosomes are potential biomarkers for SLE.

ACKNOWLEDGMENTS

We thank the staff of the Institute of Biological Chemistry of Academia Sinica (Taipei, Taiwan) for mass spectrometry. We thank the staff of the Core Facilities of the National Health Research Institutes (NHRI; Taiwan) for tissue sectioning/H&E staining and confocal microscopy. We thank the staff of the Transgenic Mouse Core of NHRI for generation of Lck-ECP–transgenic mice. We also thank the staff of the (AAALAC-accredited) Laboratory Animal Center of NHRI for mouse housing and serum chemistry assays.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Tan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design

Chuang, Tan.

Acquisition of data

Chuang, M. Chen, Y. Chen, Ciou, Hsueh, Tsai.

Analysis and interpretation of data

Chuang, M. Chen, Y. Chen, Ciou, Hsueh, Tsai, Tan.