T cell–derived exosomes induced macrophage inflammatory protein‐1α/β drive the trafficking of CD8+ T cells in oral lichen planus

Abstract Oral lichen planus (OLP) is a T cell–mediated chronic inflammatory disease with uncertain aetiology. Exosomes are nanosized particles with biological capacities. Here, we aimed to study the effects of T cell–derived exosomes (T‐exos) on the pathogenesis of OLP and its mechanism. T‐exos were incubated with Jurkat cells for 48 hours, and 26 cytokines in the supernatant were measured by luminex assay. The expression of macrophage inflammatory protein (MIP)‐1α/β was detected using immunohistochemistry and ELISA; that of CCR1/3/5 on peripheral T cells was determined by flow cytometry. Transwell assay was performed to investigate the chemotactic effect of MIP‐1α/β, and cells in the lower chambers were examinated by flow cytometry. As a result, OLP T‐exos elevated the production of MIP‐1α/β, which were highly expressed in OLP tissues and plasma. CCR1/5 were markedly expressed on OLP peripheral T cells, and the majority of CCR1/5+ T cells were CD8+ T cells. Besides, MIP‐1α/β promoted the migration of OLP mononuclear cells, while inhibiting CCR1/5 significantly decreased the trafficking of mononuclear cells, especially that of CD8+ T cells. Conclusively, OLP T‐exos‐induced MIP‐1α/β may drive the trafficking of CD8+ T cells after binding with CCR1/5 in OLP, contributing to the development of OLP.

Chemokines released by activated T cells and keratinocytes are able to elicit the migration of CD8 + T cells into the epithelium, and the followed cytotoxic CD8 + T cell response mediates the apoptosis of keratinocytes. 3,5 Chemokines mediated CD8 + T-cell recruitment and cytokines orchestrated inflammatory response are believed to be responsible for the sustained inflammation in OLP lesions. [4][5][6] Multivesicular bodies (MVBs) are unique organelles in the endocytic pathway that contain intraluminal vesicles (ILVs) in their lumen. 7 The formation of MVBs and ILVs is certainly complex, being driven by the endosomal sorting complex required for transport (ESCRT) composed of approximately 30 proteins that assemble into 4 complexes, ESCRT-0, -I, -II and -III. 8 Among them, Hrs from ESCRT-0 recruits Tsg101 of the ESCRT-I, and ESCRT-I is then involved in the recruitment of ESCRT-III via ESCRT-II or Alix. 7 In addition, MVBs and ILVs can form in absence of ESCRT, where tetraspanins such as CD63 and CD81 can be involved. [9][10][11] MVBs fusing with the plasma membrane release the ILVs, which are then called exosomes. 8 In this process, Rab27A and Rab27B are critical for the docking of MVBs at the plasma membrane, thereby promoting exosome secretion. 10 Exosomes, with a size ranging from 50 to 150 nm and 1.13 to 1.19 g/mL of density in sucrose, have a saucer-shape morphology and contain characteristic proteins such as CD63 and CD9. 12 Once being engulfed by adjacent and distant cells, exosomes can modify the overall functions of target cells, playing a critical role in autoimmune and chronic inflammatory diseases, such as rheumatoid arthritis and multiple sclerosis. 13 Recently, the biological capacity of T cell-derived exosomes (T-exos) in inflammation is increasingly gaining attention. Li et al found that exosomes originated from interleukin (IL)-12-stimulated cytotoxic T cells were able to activate bystander CD8 + T cells. 14 Wahlgren et al reported that exosomes released by activated CD3 + T cells were capable of promoting proliferation in autologous resting CD3 + T cells. 15 We have previously found that OLP plasma-derived exosomes enhanced the proliferation and migration of T cells, as well as the secretion of interferon (IFN)-γ. 16 However, the effects of human T-exos on OLP have not been elucidated yet.
In this study, we detected the expression of MVB-related proteins in OLP T cells, isolated T-exos and determined their effects on cytokine secretion. We investigated the expression and function of the chemokines which could be up-regulated by OLP T-exos and found a chemotactic effect on OLP mononuclear cells. In particular, we analysed the subtype of the migrated T cells and studied the mechanism of this chemotactic effect. Our findings illustrated the role of OLP T-exos in the inflammation of OLP, which may affect the production of cytokines and aggravate the lesional CD8 + T cells infiltration. and 47.2 ± 12.5 years old, respectively. Tissues were taken from the oral lesions of OLP patients or healthy human mucosa of controls.

| Participants and sample collection
Venous blood was drawn from OLP patients and HCs. All the patients with OLP met the criteria of inclusion and exclusion as we described previously (Appendix 1). 16 All the HCs were at least 18 years old and signed written informed consent; neither had any systemic disorders nor any other oral lesions; smokers and alcoholics were excluded. The severity of OLP was measured using reticular, atrophic, erosive (RAE) scoring system that we have proposed earlier (Table 1). 17

| Immunohistochemistry
Five µm thick formalin-fixed, paraffin-embedded tissue sections were deparaffinized and rehydrated. Antigens were retrieved with sodium citrate buffer using microwave. Sections were treated with hydrogen peroxide and blocked with normal goat serum. After incubating with anti-MIP-1α (A7568; ABclonal) and anti-MIP-1β (A1671; ABclonal) primary antibodies at 4°C overnight, samples were washed and incubated with HRP polymer conjugated secondary antibody. Immunoreactivity was visualized by diaminobenzidine solution followed by haematoxylin counterstain. The expression level was measured by integrated optical density (IOD) value using Image Pro Plus 6.0 software. The total score of all 10 areas ΣR + Σ(A × 1.5) + Σ (E × 2.0)

TA B L E 1 RAE scoring system for OLP
Note: The oral cavity of each individual was divided into 10 sites: upper/ lower labial mucosa, right buccal mucosa, left buccal mucosa, dorsal tongue, ventral tongue, floor of mouth, hard palate mucosa, soft palate/ tonsillar pillars, maxillary gingiva, mandibular gingiva.

| Quantitative real-time real-time polymerase chain reaction
The human peripheral blood mononuclear cells (PBMCs) were iso-
Jurkat cells were incubated in RPMI 1640 supplemented with 10% exosome-free foetal calf serum and activated with CD3 and CD28 for 48 hours.

| Purification of exosomes
Supernatants of human T cells were collected from 5-6 days cell cultures, and exosomes were purified as followed. Briefly, culture

| Characterization of purified exosomes
Purified exosomes suspended in PBS were dropped on a copper mesh and incubated at room temperature for 5 minutes. After fixing with 2% uranyl acetate for 3 minutes, samples were visualized using a transmission electron microscope (TEM; Tecnai G2 Spirit BioTwin, 80 kV; FEI).
The size of exosomes was identified using nanoparticle tracing assay

| Confocal microscopy
Purified T-exos were labelled with PKH67 Fluorescent Cell Linker Kits (MIDI67; Sigma-Aldrich) and then diluted with PBS and ultracentrifuged at 120 000 g for 70 minutes at 4°C to remove unbound dye.
After staining with 10 μmol/L Dil fluorescent cell membrane probe

| Enzyme-linked immunosorbent assay
The expression level of MIP-1α and MIP-1β in human plasma was

| Statistical analysis
All data were analysed by Graphpad Prism 7.0. When data were normally distributed and showed homogeneity of variance, significance of mean differences was determined by unpaired Student's t test (two groups) or one-way ANOVA with Tukey's multiple comparison test (more than two groups); otherwise, they were calculated by non-parametric Mann-Whitney U tests (two groups) or Kruskal-Wallis test (more than two groups). Spearman's correlation test was used to examine the clinical correlations. Experimental data were presented as mean ± SEM. Differences were considered statistically significant at P < 0.05.

| MVB-related genes and proteins were significantly up-regulated in OLP T cells and correlated with disease severity
To explore if there was a connection between exosomes and T cells,  Figure 1D-a), so was the gene expression of Tsg101 ( Figure 1D-e).
These data suggested that exosomes released by OLP T cells may be implicated in the pathogenesis of OLP.

| Identification and internalization of exosomes derived from human peripheral T cells
We then isolated T-exos by ultracentrifugation and identified it on the basis of morphology, size distribution and membrane composition. Nanoparticle tracing assay showed that the size of the T cellderived extracellular vesicles was homogeneously distributed with an average size of 118.7 ± 32.1 nm and a diameter peak at 117 nm.
A classic saucer-shape morphology of these vesicles was observed under the TEM. Besides, the amount of exosomal markers CD63 and CD9 was significantly higher in exosomes than in the lysate of T cells ( Figure 2A).
To determine whether T-exos could be internalized, PKH-67 la-

| Increased expression of MIP-1α/β and CCR1/5 + T cells in OLP
To determine the effects of MIP-1α/β on OLP, we first detected the expression of MIP-1α/β and their receptors CCR1/3/5. The expression of MIP-1α and MIP-1β was remarkably increased in the OLP lamina propria and plasma compared with HCs ( Figure 4A,B). In peripheral T cells, CCR1 and CCR5 were highly expressed on OLP T cells when compared with controls, whereas the expression of CCR3 showed no significance between the two groups and expressed on only 0%-0.8% T cells ( Figure 4C; Appendix 3). We further analysed the proportion of CD4 + and CD8 + cells in CCR1/5 + T cells and found that the majority of CCR1 + T cells were CD8 + T cells, and in OLP the percentage of CCR5 + CD8 + T cells was significantly higher than that of CCR5 + CD4 + T cells ( Figure 4D).

| MIP-1α/β induced the migration of OLP CD8 + T cells via CCR1 and CCR5
Then, we performed transwell assay and flow cytometry to test the chemotactic effect of MIP-1α/β on OLP T cells and its possible mechanism. The results showed that MIP-1α/β stimulation significantly enhanced the migration of OLP PBMCs, and the inhibition on CCR1/5 markedly decreased the total number of migrated PBMCs ( Figure 5A). Although MIP-1α treatments elevated the proportion of migrated CD8 + T cells in comparison with the DMSO groups, the alteration was slight ( Figure 5B; Appendix 4). Before inhibiting CCR1/5, the predominant subtype in the lower chambers was CD8 + T cells. Inhibition on the CCR1 or CCR5 significantly diminished the proportion of migrated CD8 + T cells; however, it did not alter the percentage of CD4 + T cells in the lower chambers ( Figure 5B).

| D ISCUSS I ON
Upon TCR activation, the production of exosomes from T cells would be highly increased, 18 which subsequently could be internalized by in situ or remote cells and display immunologic functions. 19 Wang et al found that exosomes derived from exhausted CD8 + T cells were able to damage the proliferation and cytokine production such as IFN-γ and IL-2 of normal CD8 + T cells, impairing the anticancer function of normal CD8 + T cells. 20 In type I diabetes, T-exos could lead to the death pancreatic β cells and increase the expression of chemokine genes, which may promote the autoimmune attack. 21 Here, we first identified the potential linkage between exosomes and OLP T cells

CO N FLI C T O F I NTE R E S T
The authors state no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request from the authors. The data that support the findings of this study are available from the corresponding author upon reasonable request.