Exosomes derived from EphB2-overexpressing bone marrow mesenchymal stem cells regulate immune balance and repair barrier function

Disruption of intestinal barrier function and an imbalance in intestinal immunity are crucial for the occurrence and development of ulcerative colitis. Because of their important roles in regulating inflammation and immunity, exosomes (Exos) released from bone marrow mesenchymal stem cells (BMSCs) may be useful for treating ulcerative colitis. The EphB/EphrinB signaling pathway plays a crucial role in the inflammatory process and the development and function of immune cells, and can mediate long-distance intercellular communication through extracellular vesicles. This study was conducted to explore the effects of pre-modified BMSC-Exos expressing EphB2 (EphB2-Exos) on immunoregulation in vitro. We transfected a lentivirus vector encoding EphB2 into BMSCs and isolated EphB2-Exos from the culture supernatant. Inflammation and oxidative damage in the human colon adenocarcinoma cell line (Caco-2) were induced by dextran sulfate sodium/hydrogen peroxide. In addition, spleen CD4+ T lymphocytes of rats were sorted in vitro. We conducted a series of experiments to explore the biological functions of EphB2-Exos. EphB2-Exos were successfully isolated and were found to significantly protect the activity, proliferation, and migration of Caco-2 cells that were inhibited by dextran sulfate sodium. EphB2-Exos alleviated inflammation and apoptosis and increased the activity of antioxidant enzymes while inhibiting oxidative stress in Caco-2 cells. EphB2-Exos restored intestinal barrier function by inhibiting the RhoA/ROCK pathway and regulated the polarization of CD4+T cells. EphB2-Exos enhanced intestinal barrier function and regulated the immune balance by inhibiting the RhoA/ROCK pathway in vitro. These findings suggest that EphB2-Exos can be applied as a cell-free therapy for ulcerative colitis.


Introduction
Ulcerative colitis (UC) is an inflammatory bowel disease that is often accompanied by alternating episodes and remissions. In recent years, the global incidence of UC has increased annually. However, existing therapeutic drugs cannot completely cure the disease and exhibit some side effects and drug resistance (Renna et al. 2014;Kaser et al. 2010). Therefore, new drugs must be developed. The pathogenesis of UC has been traditionally thought to involve genetics, environmental factors, immune disorders, and intestinal microbes; however, the precise mechanism underlying the pathogenesis of this disease remains unclear. Recently, through the accelerated pace of research, the innate immune system and its interaction with gut microbes and the adaptive immune system have been shown to be important factors in UC (Kaser et al. 2010).
Exosomes are extracellular vesicles with a doublelayer membrane structure that are derived from a wide range of sources; they are typically 30-150 nm in diameter and can carry nucleic acids, proteins, and other biologically active ingredients Xu et al. 2023). Bone marrow mesenchymal stem cells (BMSCs) are a type of multipotent stem cells that reside with hematopoietic stem cells in the bone marrow cavity and exhibit multi-differentiation abilities. Initially, these cells were used as therapeutic tools to treat bone and cartilage diseases. Multiple reports have found, BMSCs can home to the injured site by producing bioactive ingredients and antiinflammatory factors, reducing inflammation at the injured site and promoting the repair and regeneration of damaged tissues (Hofstetter et al. 2002;González et al. 2009;Jurewicz et al. 2010;Ning et al. 2022). Other recent studies have demonstrated that exosomes released from BMSCs (BMSC-Exos) can increase the production of anti-inflammatory factors, inhibit inflammatory responses, and regulate immunity. BMSC-Exos have been shown to reduce the expression of the pro-inflammatory cytokines, interleukin-1 beta (IL-1β) and tumor necrosis factor-alpha, and increase the expression of anti-inflammatory cytokine transforming growth factor β (TGF-β) in in vitro studies. In addition, BMSC-Exos regulate the proliferation and transformation of T cells, promote the transformation of Th1 cells into Th2 cells, and reduce the production of Th17 cells (Chen et al. 2016;Harrell et al. 2019;Pando et al. 2023;Shen et al. 2022). As a cell-free therapy, exosomes have low immunogenicity, a long half-life in systemic circulation, and the ability to cross the blood-brain barrier (Gyorgy et al. 2015;Kalani et al. 2014;Sun et al. 2010). Based on these advantages, exosomes may be useful for treating inflammatory bowel disease (IBD).
Culture conditions, such as genetic modification, can enhance the functions of mesenchymal stem cells (MSCs) in vivo and in vitro (Hu and Li 2018). Pre-treated MSCs show better paracrine function and homing ability to the injured site than untreated MSCs (François et al. 2012;Carrero et al. 2012). Recent studies have demonstrated that in experimental colitis mice, MSCs treated with IL-1β exhibit an increased ability to migrate to the spleen, mesenteric lymph nodes, and colon, and there is a low proportion of M1 macrophages in the abdominal cavity of the mice (Carrero et al. 2012;Fan et al. 2012). Erythropoietin-producing hepatocellular (Eph) receptors form the largest family of receptor tyrosine kinases, and bidirectional signal transduction between Eph and its ligand Ephrin activates bidirectional communication between the receptor and ligand cells (Lisabeth et al. 2013;Pasquale 2010Pasquale , 2008Kullander and Klein 2002). The Eph-Ephrin axis can control cell migration (Barquilla and Pasquale 2015;Cho et al. 2018;Himanen et al. 2007). Recent studies have revealed that intestinal epithelial cells 6 cultured in vitro show high expression of EphrinB1 and EphrinB2 ligands, and during intestinal repair, their reverse signals can regulate inflammation and accelerate intestinal injury repair (Hafner et al. 2005). Eph/ephrin molecules have been reported to be involved in the development and function of immune cells (Wu and Luo 2005). EphrinB1, which is highly expressed by human primary T cells, shows high affinity with EphB2, which is highly expressed by MSCs (Nguyen et al. 2013). Previous studies have found that EphrinB1/EphB2 reverse signaling inhibits the proliferation of mouse T cells (Kawano et al. 2012). However, the effect of the EphB2/EphrinB1 axis on BMSCs-Exos and its potential role in UC immune regulation in vitro have not been reported previously.
Caco-2 cells are a human colon epithelial cancer cell line, and Caco-2 cell monolayer is commonly used to simulate the intestinal epithelium. When grown to confluence, cell polarity and tight junctions are established in the Caco-2 cell monolayers. This model has become a standard tool for predicting intestinal permeability and drug transport in humans.
In the present study, we aimed to investigate the effect of BMSCs-Exos pretreated with lentivirus overexpressing EphB2 on the biological function of Caco-2 cells exposed to dextran sodium sulfate (DSS) or H 2 O 2 . We also evaluated the modulatory effect of EphB2-Exos on the polarization of Th1, Th17, and regulatory T (Treg) cells in vitro.

BMSC isolation, expansion, and identification
All animals used in this study were purchased from Hubei Bente Biotechnology Co., Ltd. (Wuhan, China). The entire research process was supported and supervised by the Animal Care and Use Committee of Tongji Medical College, Huazhong University of Science and Technology, China. BMSCs were isolated from male Sprague Dawley rats weighing 60-100 g according to previously described methods (Zhang et al. 2020;Xu et al. 2019;Sultana and Neelakanta 2023). Briefly, the bone marrow in the rat femur and tibia flushed out using a 5-mL sterile syringe was cultured in DMEM/F12 (Gibco, Grand Island, NY, USA) containing 10% exosome-depleted fetal bovine serum (Gibco, Australia) in a humidified atmosphere of 5% CO 2 . After culturing the cells for 48 h, the medium was changed to remove non-adherent cells. When the cells reached 80-90% confluence, they were digested with trypsin (0.05% Trypsin-EDTA, Gibco, Canada) and then subcultured and expanded. At passages 3-5 (P3-P5), the cells were labeled with antibodies against CD29, CD90, CD45, and CD11b (BioLegend, San Diego, CA, USA), and BMSCs were phenotyped by flow cytometry (Beckman Coulter Cytoflex, America).

Construction of recombinant lentivirus carrying EphB2
We used GenBank to design forward (5′-GTC AAC ACG CTG GAC AAG AT-3′) and reverse (5′-CCT TAT AGT CCT TAT CAT CGTC-3′) primers, which were synthesized by Shanghai Generay Biotech Co., Ltd. (Shanghai, China). After the amplified EphB2 gene fragment was inserted into the lentiviral vector GV342 (Ubi-MCS-SV40-puromycin), the linearized GV342 vector was used to amplify the EphB2 gene fragment using polymerase chain reaction (PCR). Positive clones with 394-bp PCR products were selected for sequencing. This recombinant lentivirus was designed with both the puromycin-resistant gc green fluorescent protein gene. After co-transfection with GV492 and packaging plasmids, pHelper 1.0 and pHelper 2.0 (Genechem), recombinant lentivirus was produced by 293 T cells.

Lentiviral transfection of BMSCs
P2-P3 BMSCs were inoculated into the culture flask and cultured for approximately 8-12 h until reaching 30-40% confluence, following which transfection was performed at a multiplicity of infection of 30-35 using an appropriate dose of HitransG P (Genechem), according to the manufacturer's instructions. The genetically engineered BMSCs used recombinant lentiviruses that simultaneously expressed EphB2 and puromycin-resistant genes. The negative control used an empty lentivirus with anti-puromycin gene but not the EphB2 gene. Approximately 48-72 h after transfection, the medium containing lentivirus was discarded and replaced with complete medium. To screen for the stable expression of EphB2-BMSCs, puromycin was added to the complete medium to a final concentration of 2-3 mg/mL, and the medium was changed daily for 7 days. The obtained stably transfected cells were expanded to generations P4-P5 and used for exosome extraction.

Exosome isolation, purification, and identification
To prepare the exosomes, we cultured P4-P5 BMSCs to approximately 80-90% confluence and then cultured the cells in serum-free DMEM/F12 for 48 h to obtain more exosomes, following which the cell culture supernatants were collected. Exosomes were acquired by ultracentrifugation, first by centrifuging the cell culture supernatant at 300 × g for 15 min and then at 2000 × g for 30 min. The obtained supernatant was centrifuged at 12,000 × g for 30 min. To remove excess cell debris, the supernatant was filtered through a 0.22-μm filter. The supernatants were ultra-centrifuged twice (Beckman, Brea, CA, USA) at 120,000 × g for 90 min, which yielded the pure exosomes. All the above steps were conducted at 4 °C ( Fig. 1).
To demonstrate the successful isolation of exosomes, the morphology of the enriched exosomes was observed using transmission electron microscopy (TEM) (Tecnai, FEI, Hillsboro, OR, USA). The marker proteins on exosomes (CD9, CD63, and TSG101) were detected by western blotting. The size distribution of exosomes was measured using nanoparticle tracking analysis (NanoSight, Salisbury, UK). After quantifying the protein concentration of exosomes using a BCA kit (Beyotime Biotechnology, China), the exosomes were stored at − 80 °C until analysis.

Cell culture
Caco-2 cells were purchased from Shanghai Zhong Qiao Xin Zhou Biotechnology Co. (Shanghai, China) and cultured in high-glucose (HG)-DMEM (Gibco, Waltham, MA, USA) containing 10% FBS, 2.5% HEPES (Sigma, St. Louis, MO, USA), and 1% antibiotics (Sigma-Aldrich, St Louis, MO, USA) and seeded onto culture flasks (25 or 75 cm 2 ). The complete medium was replaced every two days; when the cells were 90% confluent, they were digested with 0.25% trypsin-EDTA for cell subculture. This cell line is commonly used in studies of intestinal epithelial wound healing (Charrier et al. 2005).

CCK-8 assay
The effect of EphB2-Exos on the viability of Caco-2 cells was evaluated in a Cell Counting Kit-8 (CCK-8, Dojindo, Kumamoto, Japan) assay. After Caco-2 cells were seeded onto 96-well plates and cultured for 12 h, the complete medium was replaced with HG-DMEM containing 2% serum, and exosomes and EphB2-Exos were added to the medium. After 24 h of treatment, CCK-8 reagent with a final concentration of 10% was added to the 96-well plate. After incubation for 2 h at 37 °C, the amount of formazan formed was detected by measuring the absorbance at 450 nm with a reference wavelength of 655 nm using a multimode plate reader (PerkinElmer EnSpire Plate Reader, Waltham, MA, USA).

Wound-healing assay
Caco-2 cells were seeded onto a six-well plate at a density of 3 × 10 5 cells per well. After the cells reached around 90% confluence, scratch analysis was performed. The scratches in each well were created using a sterile 10-µL pipette tip. The scraped floating cells were washed twice with PBS, and the remaining cells were incubated with EphB2-Exos. An optical microscope (Olympus, Tokyo, Japan) was used to photograph the same locations along the scratch wound at 0 and 24 h. ImageJ software 8.0 (NIH, Bethesda, MD, USA) was used to measure the areas of the scratch wounds. The culture medium of Caco-2 cells grown in sixwell plates was collected to detect peroxides and antioxidant enzymes. The cells were pretreated with exosomes and EphB2-Exos for 12 h. To mimic oxidative damage, in addition to the control group, 250 µM hydrogen peroxide (H 2 O 2 , Sigma-Aldrich, St. Louis, MO) was added to the Caco-2 cells in each group. The experimental groups were as follows: (1) control, (2) H 2 O 2 (250 μmol/L), (3) exosomes (30 μg/mL), and (4) EphB2-Exos (30 μg/ mL). After treatment for 24 h, the cells were collected and used to detect the concentrations of myeloperoxidase (MPO), superoxide dismutase (SOD), glutathione (GSH), malondialdehyde (MDA), and nitric oxide (NO) using a commercial assay kit (Nanjing Jiancheng, China), and the cell culture media from each group were collected for subsequent experiments.

Measurement of intracellular reactive oxygen species levels
The reactive oxygen species (ROS) generation test was performed as previously described (Hu et al. 2016). For the experiment, 3 × 10 5 Caco-2 cells per well were seeded into a 6-well black plate in complete culture medium. The groups of cells were prepared and treated as described previously. After these steps, the medium was aspirated, and cells were washed twice with PBS. Next, 1 mL of 2′,7′-dichlorofluorescein diacetate (DCFH-DA) was added to each well at a concentration of 10 µM, and the cells were further incubated at 37 °C under 5% CO 2 for 60 min. After incubation, DCFH-DA was removed, and the cells were washed with PBS. The negative controls contained cells in DMEM, whereas positive controls contained cells in DMEM with 20 µM hydrogen donor for 60 min. The medium was aspirated, and the cells were collected with PBS. These suspensions were centrifuged at 300 × g for 5 min; the pellet was washed twice with PBS and then suspended in sterile PBS. The fluorescence of the cells was assessed by flow cytometry. The average 2′,7′-dichlorofluorescein (DCF) fluorescence was determined as a percentage compared to the negative control, which was assumed to be 100%.

Enzyme-linked immunosorbent assay
Cell culture medium samples were obtained from the different groups, and the protein levels of IL-1β, IL-6, IL-17A, IL-10, and TGF-β1 in the media were measured using enzyme-linked immunosorbent assay kits (Elabscience Biotechnology Co., Ltd., Houston, TX, USA) according to the manufacturer's instructions.
Fluorescein isothiocyanate-dextran permeability assay To measure intestinal barrier function, we seeded 5 × 10 4 Caco-2 cells onto the apical side of 24-well Transwell polyester membrane filters (6.5 mm diameter, 3 µm pore size; Corning, Inc., Corning, NY, USA) and cultured the cells for at least 21 days. Cells in the upper chamber were pretreated with exosomes or EphB2-Exos for 24 h, and 3% DSS dissolved in DMEM was added to the upper chamber for 8 h. To determine the permeability of cell monolayers, the cells were incubated with fluorescein isothiocyanatedextran (FD-4, 4 kDa, 1 mg/mL; Sigma-Aldrich, St. Louis, MO) in the upper chamber for 2 h at 37 °C; subsequently, 100 µL of samples from the outer chamber was collected for assessment using a multimode plate reader at excitation and emission wavelengths of 485 and 535 nm, respectively. The dextran concentration in each sample was calculated using a fluorescein isothiocyanate-dextran standard curve.

Western blotting
Caco-2 cell proteins were extracted using RIPA lysis solution containing a cocktail of protease inhibitors (Servicebio Technology, Wuhan, China). After treatment, the Caco-2 cells were washed twice with cold PBS, and RIPA lysis solution was added to the cells. The protein concentration was determined using a BCA protein assay kit (Beyotime Biotechnology, China). The same mass of protein was separated using 8-12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene fluoride membranes (Millipore, Billerica, MA, USA). After blocking with 5% skimmed milk, the membranes were incubated with primary antibodies against EphB2 ( and β-actin (1:10,000; CST) overnight at 4 °C. Next, the membranes were washed three times with PBS containing Tween 20 and incubated with the appropriate secondary antibody at 25 °C for 1 h. Finally, the membranes were visualized according to the manufacturer's instructions.

Statistical analyses
All experimental data are presented as the mean ± standard error of the mean. Comparisons between multiple groups were evaluated using one-way analysis of variance followed by Tukey's post hoc test. Statistical analysis was performed using GraphPad Prism (GraphPad Prism 8.4.2, USA). P < 0.05 was considered statistically significant.

Characterization of BMSCs and exosomes
BMSCs were purified from Sprague Dawley rat bone marrow tissue and characterized based on their morphological properties and immunophenotypes at passage 3. Spindle-shaped BMSCs exhibited plastic adhesion and formed swirling colonies ( Fig. 2A). The purity of the BMSCs was evaluated by flow cytometry analysis of specific cell surface markers. As illustrated in Fig. 2B, BMSCs highly expressed CD29 and CD90 but expressed minimal levels of CD11b and CD45. The former are surface markers of hematopoietic cells, whereas the latter are surface markers for bone marrow progenitor cells. The extracted exosomes were characterized using TEM, nanoparticle tracking analysis, and western blotting. TEM images showed that most exosomes were within 200 nm in diameter and had a cup-or round-shaped morphology (Fig. 2C). The diameter of the exosomes was approximately 85 nm (Fig. 2E). The expression of TSG101, CD63, and CD9 proteins was detected by western blot analysis (Fig. 2D). Collectively, these data confirmed the successful isolation of exosomes from BMSCs.

Expression of EphB2 was upregulated in BMSCs and exosomes after lentivirus intervention
We successfully constructed a lentivirus carrying the EphB2 gene and obtained BMSCs stably overexpressing EphB2 by in vitro transduction with lentivirus and puromycin screening. EphB2-BMSCs with gcGFP gene were observed to emit green fluorescence under a fluorescence microscope at 72 h of lentivirus transfection (Fig. 2F). The corresponding reverse transcription-PCR results were obtained from our previous study (Yu et al. 2021). The expression of EphB2 was upregulated successfully in BMSCs and

Fig. 3 EphB2-Exos enhanced the viability and migration of Caco-2 cells.
A The morphological image of Caco-2 cells observed under optical microscope (Original magnification, × 40). B CCK8 assay was performed to evaluate Caco-2 cells viability after treatment with DSS and EphB2-Exos. The cell viability was normalized to control group. C Wound healing assay of the Caco-2 cells. D Quantitative analysis of gap areas and wound closure in Caco-2 cells. Statistical analysis performed using one-way ANOVA with a Tukey Kramer post-hoc test. ** P<0.01, *** exosomes after lentivirus intervention compared to the negative groups.

EphB2-Exos promote the viability and migration of Caco-2 cells
To explore the influence of EphB2-Exos on the viability and migration of Caco-2 cells (Fig. 3A), we defined 4 groups (control, DSS, exosome, and EphB2-Exos groups). The viability of Caco-2 cells treated with EphB2-Exos was measured using the CCK-8 assay, which showed that DSS inhibited the activity of Caco-2 cells, whereas exosomes protected against this effect, and the protective effect of EphB2-Exos was stronger (Fig. 3B). To explore the effect of EphB2-Exos on injury repair in vitro, we conducted a wound-healing assay. DSS inhibited damage repair, whereas exosomes promoted wound healing; among the treatments, EphB2-Exos exhibited the strongest protective effect (Fig. 3C, D). These results indicate that EphB2-Exos rescued the viability of Caco-2 cells inhibited by DSS and promoted wound healing.

EphB2-Exos inhibit oxidative stress in Caco-2 cells
In this experiment, to explore whether EphB2-Exos could affect the oxidative stress of Caco-2 cells, we exposed the cells to H 2 O 2 and constructed an oxidative damage model of Caco-2 cells. The enzyme activities of superoxide dismutase and glutathione in the cell lysate and supernatant were significantly decreased in the H 2 O 2 group. The malondialdehyde and myeloperoxidase levels in the cell lysate or supernatant increased after adding H 2 O 2 . The protective effect of superoxide dismutase against glutathione was observed in cells treated with exosomes, particularly in the EphB2-Exos group (Fig. 4A). These results indicated that EphB2-Exos alleviated the oxidative stress response of Caco-2 cells caused by H 2 O 2 .

Measurement of intracellular ROS production
As ROS production is common in every cell line and is typically involved in inflammatory conditions, we analyzed redox signaling by measuring ROS production. We observed significant increases in ROS levels in Caco-2 cells. Intracellular ROS levels were analyzed using the fluorescent probe 6-carboxy 2′,7′-dichlorodihydrofluorescein diacetate (carboxy-H2DCFDA). Carboxy-H2DCFDA diffuses into cells and is hydrolyzed to DCFH. DCFH trapped within the cells is converted to DCF by ROS-mediated oxidation. The H 2 O 2 group exhibited a higher level of ROS than the control group. Although the exosomes significantly reduced ROS production compared to H 2 O 2 , the effect of EphB2-Exos was stronger (Fig. 4B). These results suggested that antioxidant enzyme activities are upregulated in Caco-2 cells.

EphB2-Exos alleviate inflammatory responses in Caco-2 cells
The concentrations of pro-and anti-inflammatory cytokines in Caco-2 cells were determined using commercially available enzyme-linked immunosorbent assay kits. The protein expression of pro-inflammatory cytokines, including IL-1β, IL-6, and IL-17A, was significantly elevated after 8 h of incubation with DSS; however, the protein levels of IL-10 and TGF-β1 were downregulated (Fig. 4C). The exosomes prevented increases in the levels of the above-mentioned pro-inflammatory cytokines. Exosomes upregulated IL-10 and TGF-β1 protein levels. Compared with exosomes, EphB2-Exos exhibited a similar but more powerful regulatory effect.

EphB2-Exos improve intestinal permeability in
Caco-2 cells FD-4 cannot pass through a complete intestinal barrier; however, FD-4 can penetrate the blood through an injured or damaged intestine. In this study, intestinal barrier function was assessed indirectly through intestinal permeability analysis. DSS disrupted the Caco-2 cell monolayer, resulting in a significant increase in the level of FD-4 in the medium, and these increases were reversed when exosomes were added. In addition, EphB2-Exos significantly reduced the levels of FD-4 in the medium (Fig. 5A).
In addition, the protein expression of ZO-1 and occludin in Caco-2 cells decreased after DSS intervention, whereas exosomes reversed this effect (Fig. 5B). These results suggest that exosomes can ameliorate DSS-induced epithelial barrier damage; this effect was more pronounced when EphB2-Exos were used. Fig. 4 EphB2-Exos alleviated inflammation and oxidative stress in Caco-2 cells. A Expression levels of MPO, MDA, GSH, and SOD in Caco-2 cells after treatment with H 2 O 2 and EphB2-Exos. B Caco-2 cells exposed to H 2 O 2 were treated with EphB2-Exos for 12 h. The intracellular ROS production was detected with ROS assay kit. DCFH-DA was added to Caco-2 cells and left for 60 min, fluorescence of oxidized DCF was followed by emission and excitation wavelengths of 535 nm and 485 nm, respectively. C Protein levels of IL-1β, IL-6, IL-17A, IL-10 and TGF-β in Caco-2 cells from the four groups by ELISA. Statistical analysis performed using one-way ANOVA with a Tukey Kramer post-hoc test. * P<0.05, ** P<0.01, *** EphB2-Exos enhance ZO-1 and occludin expression by inhibiting the RhoA-ROCK signaling pathway To explore whether EphB2-Exos improves intestinal permeability by regulating the tight junction protein, the expression of ZO-1 and occludin in Caco-2 cells was measured by western blotting. After DSS treatment, the expression levels of ZO-1 and occludin EphB2-Exos inhibited RhoA/ROCK pathway in Caco-2 cells. A Western blotting analysis of EphB2 and ephrin-B1 proteins and quantitative analysis of EphB2 and ephrin-B1 protein levels in Caco-2 cells. The protein expression was normalized to β-actin. B Western blotting analysis of RhoA, ROCK1 and ROCK2 proteins and quantitative analysis of RhoA, ROCK1 and ROCK2 protein levels in Caco-2 cells. The protein expression was normalized to β-actin. Statistical analysis performed using one-way ANOVA with a Tukey Kramer post-hoc test. * P<0.05, ** P<0.01, *** proteins decreased, and exosomes reversed this change (Fig. 5B). The RhoA-ROCK signaling pathway is involved in regulating T cell activation and intestinal barrier damage (Bros et al. 2019;Zhao et al. 2020). Therefore, we analyzed the protein expression of RhoA, ROCK1, and ROCK2. ROCK1 and ROCK2 protein levels increased after DSS intervention in Caco-2 cells and decreased significantly after treatment with EphB2-Exos (Fig. 6A, B). Activation of the RhoA/ROCK signaling pathway was significantly induced after DSS incubation, and EphB2-Exos significantly inhibited the activation of this signaling pathway. These results suggest that EphB2-Exos enhance the expression of ZO-1 and occludin by inhibiting the RhoA/ROCK signaling pathway.

Effect of EphB2-Exos on the polarization of Th1, Th17, and Treg cells in vitro
To determine the effect of EphB2-Exos on the polarization of CD4 + T cells, rat spleen CD4 + T lymphocytes were obtained by magnetic bead sorting (Fig. 7A). After 72 h of stimulation with exosomes and EphB2-Exos and corresponding cytokines (Fig. 7A), we assessed the proportion of Th1, Th17, and Treg cells among all CD4 + T cells. After treatment with the exosomes, the proportions of CD4 + IFN-γ + Th1 cells and CD4 + IL17A + Th17 cells were significantly decreased (Fig. 7B, C); in contrast, the proportion of CD4 + Foxp3 + Treg cells increased (Fig. 7D). After EphB2-Exos treatment, each subgroup showed the same change as the exosome group; however, the change was more exaggerated. These results indicate that EphB2-Exos inhibited CD4 + T lymphocyte differentiation into Th1 and Th17 cells but induced them to differentiate into Treg cells in vitro (Fig. 7E-G).

Discussion
We found that EphB2-Exos effectively reversed the inhibitory effects of DSS on Caco-2 cell viability and migration. In vitro experiments showed that EphB2-Exos alleviated the inflammatory response and oxidative stress of Caco-2 cells and enhanced intestinal barrier repair by inhibiting the RhoA-ROCK pathway. Additionally, EphB2-Exos were found to regulate the polarization of Th1, Th17, and Treg cells in vitro.
These findings highlight that EphB2-Exos can relieve intestinal inflammation, repair the intestinal barrier, and restore intestinal immune balance, demonstrating the potential of EphB2-Exos for treating UC.
Innate immunity (mainly impaired intestinal barrier and microbial interaction) and adaptive immune disorders have been generally thought to play an essential role in the occurrence and development of UC (Kaser et al. 2010;Geremia et al. 2014). Intestinal barrier damage leads to increased intestinal permeability, which may result from regulatory defects of tight junction proteins. Intestinal epithelial cells have the basic function as a physical barrier, in addition to producing antimicrobial peptides and other substances to build defenses and prevent the invasion of bacteria and other microorganisms (Mehandru and Colombel 2021;Vancamelbeke and Vermeire 2017;Odenwald and Turner 2017). Studies have shown that patients with IBD are characterized by intestinal epithelial barrier damage and dysfunction, and healthy family members of these patients also have intestinal epithelial barrier dysfunction without IBD, suggesting that these factors cause IBD (May et al. 1993;Katz et al. 1989). We previously showed that EphB2-EVs can alleviate intestinal inflammation in rats with DSS-induced experimental colitis and increase the expression of intestinal tight junction proteins (Yu et al. 2021). In this study, we demonstrated that EphB2-Exos can enhance the function of Caco-2 cells, improve intestinal permeability, and promote barrier repair.
Recent studies have focused on modifying exosomes to enhance their therapeutic effects. Eph/ ephrin molecules participate in many important physiological activities in the body, including embryogenesis, inflammation regulation, and blood vessel formation. In addition, they regulate the adhesion and migration of MSCs. Wnt/β-catenin and its downstream EphB/EphrinB signal transduction are key to the proliferation and localization of intestinal epithelial cells during migration (Batlle et al. 2002). After the intestinal epithelium is damaged, rapid recovery of the intestinal surface is highly dependent on the ability of epithelial cells on the edge of the damage to rapidly form pseudopodia, reorganize the cytoskeleton, and quickly migrate to the damaged site (Jacinto et al. 2001;Wilson and Gibson 1997). To protect the host from several pathogenic microorganisms and exposure to various intestinal pathogens and toxic factors, the rapid healing ability of the intestinal epithelium is particularly important (Wilson and Gibson 1997). Recent studies have shown that EphB/ ephrinB signaling plays a key role in the repair and maintenance of intestinal tissues (Batlle et al. 2002). In the present study, we administered EphB2-Exos to Caco-2 cells treated with DSS. ZO-1 and occludin protein expression indicated that EphB2-Exos promoted the expression of tight junction proteins and repaired the intestinal tract. In addition, FD-4 analysis showed that DSS destroyed the permeability of the cell monolayer; EphB2-Exos reversed this effect, indirectly indicating that EphB2-Exos can enhance intestinal permeability and repair the intestinal epithelial barrier.
RhoA is a small GTPase in the Rho family and is involved in several cellular processes, including cell proliferation and differentiation. Rho-related kinases (ROCK), including ROCK1 and ROCK2, have been identified as key effectors of RhoA, which regulates various physiological functions through the phosphorylation of downstream targets such as MYPT-1, MLC, and ERM. The signaling pathway mediated by Rho GTPase plays an important role in regulating the immune response mediated by T cells, including the development, activation, and differentiation of T cells. Recently, studies have found that ROCK activity is elevated in various autoimmune diseases, including systemic lupus erythematosus, rheumatoid arthritis, and idiopathic pulmonary fibrosis. In addition, an increase in RhoA-ROCK related pathway protein levels has been observed in the intestinal mucosa of patients with active CD (Vancamelbeke and Vermeire 2017). In mice and humans with arthritis, ROCK2 has been shown to upregulate IL-17A and IL-21 expression in CD4 + T cells by phosphorylating IRF and STAT3 (Yoseph et al. 2016). ZO-1 and occludin are essential proteins for maintaining the permeability of the intestinal epithelial barrier, and Rho kinase regulates tight junctions, specifically occludin. Owing to the action of DSS, the cells release numerous inflammatory factors; the Ras homolog (Rho) protein is activated, and it binds to ROCK. The activated RhoA/ROCK signaling pathway regulates polymerization of the actin cytoskeleton, thereby increasing the distance between intestinal epithelial cells and regulating intestinal permeability (Jin and Blikslager 2020;Wang et al. 2018). Our results showed that DSS treatment leads to decreased levels of the tight junction proteins, ZO-1 and occludin, in Caco-2 cells and increased levels of RhoA, ROCK1, and ROCK2; following EphB2-Exos treatment, the levels of tight junction proteins increased. After this reversal, the levels of constituent proteins in the RhoA/ROCK signaling pathway decreased. These findings indicate that the RhoA/ROCK signal transduction is involved in repairing the intestinal epithelial barrier by EphB2-Exos.
A common feature of patients with IBD is the infiltration of inflammatory CD4 + T cells. Among these cells, the main pro-inflammatory cells are Th1 and Th17; their infiltration is accompanied by a lack of immunosuppressive T cells, such as Treg cells that express Foxp3 (Tindemans et al. 2020;Liu et al. 2019;Rovedatti et al. 2009). EphB/ephrin B signaling has an important influence on the development of the thymus and T-cell proliferation. We found that mice lacking the EphB2 gene showed insufficient thymic lymphocytes and decreased cell migration. The bidirectional conduction between EphB/ ephrin-B is essential for MSC-mediated suppression of activated T cells (Nguyen et al. 2013). In previous studies, we investigated the effect of EphB2-EVs on Th17 and Treg cells in the spleen and mesenteric lymph nodes of experimental colitis rats. Here, we examined the effect of EphB2-Exos on the polarization of Th1, Th17, and Treg cells in vitro by isolating CD4 + T lymphocytes from the rat spleen. The results indicate that EphB2-Exos effectively prevent the differentiation of CD4 + T cells into Th1 and Th17 cells and significantly increase the proportion of Treg cells in CD4 + T cells. Additionally, these data demonstrate that EphB2-Exos can regulate the balance between Th1 and Th17/Treg cells. The present study provides new evidence regarding the effect of EphB2-Exos on improving intestinal barrier dysfunction and regulating intestinal immune balance. It additionally indicates the potential of a new treatment strategy for UC and other immune-related diseases.

Conclusions
In conclusion, exosomes released from EphB2-BMSCs enhanced the viability of Caco-2 cells and promoted their proliferation and migration, thereby promoting barrier repair and regulating T cell polarization (Fig. 8). Intestinal barrier repair was achieved by inhibiting the RhoA/ROCK pathway. However, further research is needed to explore whether EphB2-Exos regulate T cell balance in vitro through the RhoA/ROCK pathway. We demonstrate a promising alternative strategy for treating UC and other immunological diseases.