Introduction

Autoimmune hepatitis (AIH) is one of the most common causes of life-threatening T-cell-mediated liver damage in humans [1]. AIH triggers T-cell-mediated autoimmune responses against liver autoantigens. CD4+ and CD8+ T cells are the major T lymphocytes that infiltrate the liver [1, 2]. CD4+ T cells are considered to be the key cells in the occurrence and development of AIH, and liver damage in AIH may be caused and mediated by CD4 T lymphocytes that recognize self-antigens [3].Presentation of peptides contained by HLA class II molecules by antigen-presenting cells to naive CD4 helper T cells is the initiation of pathogenesis [4,5,6].

AIH remains a serious health challenge worldwide because it can eventually result in liver cirrhosis, liver cancer, and even death [6,7,8]. AIH treatment remains a clinical issue that cannot be ignored. Therefore, finding a potential therapeutic strategy is urgent for the treatment of acute hepatitis.

Mesenchymal stromal cells (MSCs) have the potential to self-renew, differentiate into multiple cell lineages and modulate immune properties [9, 10]. MSCs come from a wide range of sources, repair tissue damage and are thought to have great potential in cell therapy [11,12,13]. MSCs can accumulate at sites of inflammation and play immunomodulatory roles [14]. In liver disease, MSCs play multiple roles in liver regeneration by inhibiting hepatocyte apoptosis, reversing liver fibrosis and reducing inflammation [15], and several studies have used MSCs to treat AIH [16, 17]. However, the curative effect of using MSCs alone still needs to be improved, and it is necessary to seek ways to enhance the immune regulatory function of MSCs [18,19,20].

Fcγ receptors play a key role in controlling innate and humoral immunity by activating the effector function of antibodies. Fcγ receptor II b (FcγRIIb, CD32b) is a low-affinity Fcγ receptor and is the only inhibitory Fcγ receptor expressed on a variety of immune cells, including B cells, macrophages, dendritic cells (DCs) and granulocytes [21].

Recent studies have shown that CD32b is also present and functional on T cells, and soluble fibrinogen-like protein 2 (sFgl2) is a key functional ligand of CD32b [22]. sFgl2 is a key immunosuppressive factor secreted by regulatory T cells (Tregs) and an important effector molecule of Tregs [23]. sFgl2 has a significant immunosuppressive effect and has been widely used in the study of organ transplantation, inflammatory bowel disease, sepsis, tumors, and parasitic diseases [24, 25].

CD32b is the only receptor of sFgl2 [26], and this study revealed that CD32b expression on CD4+ T cells could be induced by coculture with MSCs. Therefore, we aimed to examine the potential of sFgl2 gene-modified MSCs in the treatment of experimental AIH, which may provide a new strategy for the clinical treatment of AIH.

Methods

Animals

Healthy adult male C57BL/6 J mice weighing 20–21 g were used in this study. The animals were housed in wire-bottomed, wire-lidded cages, allowed access to chow and water ad libitum and were acclimatized for 1 week before experiments in a temperature-controlled room (25 ± 2 °C) with a 12-h light/dark cycle. All animal experiments were performed according to the Chinese Council on Animal Care guidelines and the basis of protocols approved by the Animal Ethical and Welfare Committee of Tianjin Medical University General Hospital (Tianjin, China) (No. IRB2021-DW-41).

Isolation of MSCs

Fat was isolated from the groins of 4-week-old C57BL/6 J mice on a clean bench and transferred to 15 mL centrifuge tubes (Corning, New York, USA) containing 5 mL of DMEM/F12 medium (Gibco, MA, USA). Collagenase type I (1 mg/mL) (Solarbio Life Science, Beijing, China) was added and incubated at 37 °C for 1.5 h under shock conditions (200 rpm). Fully digested adipose tissue was filtered through a 40-μm mesh strainer and centrifuged at 1500 rpm for 5 min to obtain cells, and this process was repeated once. The cells were resuspended in DMEM/F12 medium (Corning, New York, USA) containing 1% antibiotic cocktail (Solarbio Life Sciences, Beijing, China) and 15% fetal bovine serum, seeded on 6 cm dishes (Corning, New York, USA) and incubated in a 37 °C incubator with 5% CO2. The medium was changed every 48 h. The isolated MSCs were used in subsequent experiments.

Transfection of sFgl2 into MSCs

Second-generation MSCs were seeded in 6-well plates at a density of 3.0 × 105/well and cultured until the fusion rate reached 60%. Then, a 1-mL transfection system composed of 40 μL of enhanced infection solution, 10 μL of lentiviral vectors carrying the Fgl2 gene and encoding green fluorescent protein (Ubi-MCS-3FLAG-CBh-gcGFP-IRES-puromycin, 5.0 × 108 TU/mL) (Genchem, Shanghai, China), and 950 μL of DMEM/F12 medium containing 15% fetal bovine serum was prepared for viral transfection. After 12 h at 37 °C, the transfection medium was replaced with complete medium. Three days later, the stably transfected MSCs (sFgl2-MSCs) were screened with 2 μg/mL puromycin (Solarbio Life Science, Beijing, China). The cells and cell supernatants were used for subsequent experiments.

Coculture of CD4+ T cells with MSCs and design of experimental groups in vitro

Naïve CD4+ T cells were isolated from C57BL/6 mouse splenocytes by a CD4+ microbead kit (Miltenyi, Germany). CD4+ T cells (3.0 × 105) were plated on 96-well plates that were precoated with CD3 (5 µg/mL) (eBioscience Inc. USA) and incubated for 2 h in a 100 µL system, and CD28 (1 µg/mL) (eBioscience Inc. USA) and IL2 (15 ng/mL) (eBioscience Inc. USA) were added. CD4+ T cells were incubated at 37 °C in CO2 for 12 h to obtain activated CD4+ T cells [27, 28]. Transfected or untransfected MSCs were collected and resuspended. Then, we cocultured activated CD4+ T cells (3.0 × 105) with MSCs (3.0 × 104) in a 48-well plate. After 72 h [29, 30], CD4+ T cells were collected in different tubes for flow cytometric analysis to detect the proportions of CD4+ T-cell subsets. The experiment included 6 groups: the untreated group, wild-type MSC (WT-MSC) group, negative control MSC (NC-MSC) group, sFgl2-MSC group, sFgl2-MSC + anti-CD32b group and recombinant Fgl2 (rFgl2) group.

The control group contained only stimulated CD4+ T cells, the rFgl2 group was supplemented with rFgl2 (600 ng/mL, R&D Systems, USA) plus stimulated CD4+ T cells, and the anti-CD32b group was supplemented with anti-CD32b antibody (15 µg/mL, BioXcell, USA) plus sFgl2-MSCs. The concentrations of rFgl2 and anti-CD32b antibodies were obtained from preliminary experiments.

Con A-induced model and experimental groups in vivo

Caudal intravenous injection of Con A has been used to induce experimental AIH to generate models [31,32,33]. Liver damage was induced by injecting Con A (Solarbio Science and Technology Co, Ltd, China) via the tail vein at a dose of 15 mg/kg [34, 35] dissolved in 150 µL phosphate-buffered saline (PBS). Thirty minutes after the administration of Con A, 30 mice were randomly assigned to the following five groups (n = 6): (i) healthy mice injected with 150 µL PBS (without Con A) via the tail vein; (ii) PBS-treated mice injected with 150 µL PBS via the tail vein; (iii) WT-MSC-treated mice intravenously injected with MSCs (1 × 106 cells/mouse) suspended in 150 µL of PBS; (iv) NC-MSC-treated mice intravenously injected with MSCs (1 × 106 cells/mouse) suspended in 150 µL PBS; and (v) sFgl2-MSC-treated mice intravenously injected with sFgl2-MSCs (1 × 106 cells/mouse) in the same manner. All mice were euthanized 24 h after the administration of Con A. Animals were euthanized by carbon dioxide (CO2) inhalation. CO2 (100%) was delivered into an enclosed space at a replacement rate of 40% per minute until the mice lost their spontaneous respiration ability and died. Blood, livers, and spleens were collected and stored for further analysis.

Flow cytometry

Antibody staining and flow cytometry were performed as described previously [36]. Third-generation MSCs were collected and stained with anti-CD29 (fluorescein isothiocyanate, FITC), anti-CD44 (phycoerythrin, PE), anti-CD45 (PE), anti-CD34 (allophycocyanin, APC), and anti-SCA-1 (APC).

CD4+ T cells were collected from in vitro cocultures, resuspended and stained with fluorescent antibodies, including anti-CD4 (FITC), anti-CD25 (PE), anti-Foxp3 (APC), anti-IL-17A (PerCP-Cyanine 5.5), anti-interferon (IFN-γ, PE), and anti-CD32b (PE, eBioscience, USA).

The spleens were collected, prepared into single-cell suspensions and used for staining. The mouse liver was flushed with 0.9% normal saline via the inferior vena cava to remove residual blood and prevent blood lymphocytes from interfering with the results of subsequent experiments. After the mouse liver was ground, a single-cell suspension was obtained through a 200-mesh filter. Hepatic lymphocytes were isolated by Percoll gradient separation (top Percoll layer: 40%; bottom Percoll layer: 70%) and divided into different test tubes for staining [37]. The fluorescent antibodies used included anti-CD3 (FITC), anti-CD4 (APC), anti-CD8 (PerCP-Cyanine 5.5), anti-CD4 (FITC), anti-CD25 (PE), anti-IL-17A (PerCP/Cyanine 5.5), and anti-Foxp3 (APC). All cell staining procedures were based on the manufacturer’s instructions.

To exclude nonspecific staining caused by dead cells, a distinction needed to be made between living cells and dead cells. Viable cells were identified with a zombie dye (APC-Cyanine 7). All antibodies for flow cytometry were purchased from BioLegend (USA) unless otherwise stated. The proportions of various phenotypes of immune cells were analyzed using FlowJo software.

MSC tracking in vivo

To determine whether MSCs could reach the site of inflammation, we used live organ imaging. MSCs were stained in vitro using CM-Dil (Ex/Em = 553/570 nm, Thermo Fisher, USA) according to the manufacturer’s instructions. Then MSCs were transplanted into mice through the tail vein 30 min after the injection of Con A. After 12 h, the liver and lung were harvested and collected fluorescence images.

Measurement of alanine aminotransferase (ALT) and aspartate aminotransferase (AST)

The serum activities of ALT and AST were used to evaluate liver function and liver damage and were measured using ALT and AST assay kits (Nanjing Jiancheng, Nanjing, China) according to the manufacturer’s instructions.

H&E staining

The liver was flushed with 0.9% normal saline in the inferior vena cava to prevent blood residue and then harvested. The liver tissues were fixed with 10% formalin, embedded in paraffin and cut into 5 μm sections for H&E staining. Then, the degree of liver injury was blindly evaluated by two pathologists based on the liver Knodell scoring system, and the area of necrosis in liver tissue was measured using iViewer.

Western blotting

Total protein was extracted from CD4+ T cells in the coculture system using RIPA lysis buffer (Solarbio, Beijing, China) containing protease inhibitors (APExBIO Technology LLC, USA). The protein samples were diluted to the same final total protein concentration and then subjected to 10% SDS‒PAGE; the proteins were transferred to polyvinylidene fluoride membranes (Merck Millipore, USA). Monoclonal antibodies for detecting SHP2 (1:1000) (Affinity, China), phospho-SHP2 (1:1000) (Affinity, China), SMAD2/3 (1:1000) (Cell Signaling Technology, USA), phospho-SMAD2/3 (1:1000) (Cell Signaling Technology, USA), GAPDH (1:1000) (Affinity, China) and β-actin (1:1000) (Affinity, China) were added to the membranes and incubated at 4 °C overnight. The membranes were washed and incubated with the corresponding diluted secondary horseradish peroxidase (HRP)-marked antibodies (1:2000) (Affinity, China) at room temperature for 1 h. Finally, the blots were developed with ECL solution and photographed with an exposure instrument.

Immunohistochemical staining

T-bet, RORγ and Foxp3 in liver tissues were examined with a streptavidin–biotin peroxidase complex-based immunohistochemical technique. The liver tissues were fixed with 10% formalin, embedded in paraffin and cut into 5-μm sections for staining. The liver sections were incubated with the following primary antibodies: anti-T-bet (1:1000, Affinity, China), anti-Foxp3 (1:1000, Affinity, China) and anti-RORγ (1:1000, Affinity, China). Under a light microscope, the numbers of T-bet-, RORγ-, and Foxp3-positive cells in 5 different fields (400 ×) in each section were counted.

Enzyme-linked immunosorbent assay (ELISA)

To determine whether sFgl2 could be secreted by sFgl2-MSCs, we measured sFgl2 in the supernatant of sFgl2-MSCs using an ELISA kit (Uscn Life Science Inc., Wuhan, China).

To determine the levels of inflammatory cytokines in each group, blood samples were collected and centrifuged at 4 °C and 1000 × g for 20 min. The levels of interleukin 10 (IL-10), interleukin 6 (IL-6), interleukin 1β (IL-1β), and tumor necrosis factor-α (TNF-α) in the serum were measured using an ELISA kit (ABclonal, Wuhan, China) according to the manufacturer’s protocol.

Statistical analysis

The data are expressed as the mean ± standard deviation (mean ± SEM). Statistical analyses were performed with SPSS 21.0 (IBM SPSS, Chicago, USA). One-way analysis of variance (ANOVA) was used for group comparisons. The results were considered significant when P < 0.05.

Results

MSCs induce the expression of CD32b on CD4+ T cells

Adipose MSCs were obtained from C57BL/6 J mice and exhibited fibroblast-like morphology under a cystoscope (Fig. 1A). Flow cytometry showed that the cells highly expressed CD29, SCA-1 and CD44 but expressed low levels of CD45 and CD34 and could be identified as MSCs (Fig. 1B). After coculture with CD4+ T cells, the results showed that CD32b expression on CD4+ T cells was significantly increased compared with that in the control group (Fig. 1E and F, P < 0.001).

Fig. 1
figure 1

MSCs induced the expression of CD32b on CD4+ T cells. A Images showing the morphology of mouse-derived MSCs (magnification 40 ×). B Flow cytometric analysis of the MSC surface markers CD29, CD34, CD44, CD45, and SCA-1. C Flow cytometric analysis of the GFP-positive rates of MSCs before and after puromycin screening. D ELISA of sFgl2 concentrations (ng/mL) in the supernatants of different kinds of MSCs. E Flow cytometric analysis of CD32b expression levels on CD4+ T cells cocultured with modified or unmodified MSCs, n = 3. F Proportions of CD32b-positive CD4+ T cells in each group. The differences between groups were calculated using ANOVA. ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs, GFP, green fluorescent protein

Full size image

CD32b is known to be the only receptor of sFgl2 [26]. Based on the known immunosuppressive function of sFgl2 [23] and MSCs [14], sFgl2 gene-modified MSCs (sFgl2-MSCs) were constructed by lentiviral transfection, and high-purity sFgl2-MSCs were obtained by puromycin screening (Fig. 1C). To determine the expression of sFgl2 in MSCs, ELISA was used to analyze the cell supernatants, and the results showed that sFgl2-MSCs successfully overexpressed sFgl2, while sFgl2 was not expressed in WT-MSCs (Fig. 1D, P  < 0.001) or NC-MSCs (Fig. 1D, P  < 0.001). After coculture with sFgl2-MSCs, the results showed that the expression of CD32b on CD4+ T cells was also increased (Fig. 1E and F, P  < 0.001), and there was no significant difference compared with that in the NC-MSCs (Fig. 1E and F, P < 0.001) and WT-MSCs (Fig. 1E and F, P < 0.001). Therefore, MSCs can induce the expression of CD32b on CD4+ T cells, and this outcome is not affected by viral transfection. In addition, we obtained MSCs with high expression of sFgl2, which laid the foundation for the following experiments.

sFgl2-MSCs increase the proportion of Tregs and decrease the proportion of Th1 and Th17 cells in CD4+ T cells by activating SHP2 and SMAD2/3 in vitro

After determining that MSCs can increase the expression of CD32b on CD4+ T cells, we examined the immunomodulatory function of sFgl2-MSCs by determining the proportion of CD4+ T-cell subsets in vitro. After coculture in vitro, the proportion of cells was determined by flow cytometry. The results showed that when only MSCs were present, the proportion of Tregs was higher than that in the control group (Fig. 2A, D, P  < 0.001), while the proportions of Th17 cells (Fig. 2B and E, P  < 0.001) and Th1 cells (Fig. 2C and F, P  < 0.001) were lower.

Fig. 2
figure 2

In vitro, sFgl2-MSCs increased the proportion of Tregs and reduced the proportion of Th17 and Th1 cells. A Dot plots showing Tregs (CD4+CD25+Foxp3+) among CD4+ T cells after coculture for 72 h. B Dot plots showing Th17 cells among CD4+ T cells after coculture for 72 h. C Dot plots showing Th1 cells among CD4+ T cells after coculture for 72 h. D Proportions of Tregs (CD4+CD25+Foxp3+) in each group, n = 3, #: rFgl2 versus control, P < 0.05. E Proportions of Th17 cells (CD4+IL-17A+) in each group, n = 3, #: rFgl2 versus control, P < 0.001. F Proportions of Th1 cells (CD4+IFN-γ+) in each group, n = 3, #: rFgl2 versus control, P < 0.001. G Levels of p-SHP2 in each group. H Levels of p-SMAD2/3 in each group. Blots were cropped, and the original blots are presented in Additional file 1: Fig. S1 and Additional file 2: Fig. S2. The differences between groups were calculated using ANOVA. ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs

Full size image

When only rFgl2 was present, the proportion of Tregs was increased compared with that in the control group (Fig. 2A and D, P  < 0.05), while the proportions of Th17 cells (Fig. 2B and E, P  < 0.001) and Th1 cells (Fig. 2C and F, P < 0.001) were decreased. Compared with that in the WT-MSC group, the proportion of Tregs in the sFgl2-MSC group containing MSCs and sFgl2 was further increased (Fig. 2A and D, P < 0.001), while the proportions of Th17 cells (Fig. 2B and E, P  < 0.001) and Th1 cells (Fig. 2C and F, P < 0.001) were further decreased. Furthermore, when CD32b was blocked, sFgl2 did not exert its effect, leading to the promotion of Treg differentiation (Fig. 2A and D, P < 0.001), and the inhibitory effects of Th17 cells (Fig. 2B and E, P < 0.001) and Th1 cells (Fig. 2C and F, P < 0.001) were less pronounced than those in the sFgl2-MSC group.

To clarify the possible pathways by which sFgl2-MSCs regulate the differentiation of CD4+ T-cell subsets, we first detected the phosphorylation level of SHP2, an important downstream target of CD32b, and the results showed (Fig. 2G) that p-SHP2 level was highest in CD4+ T cells in the sFgl2-MSC coculture group, followed by the rFgl2 treatment group; and p-SHP2 level was lower in the groups with MSCs alone or CD32b blockade. Since SMAD2/3 is downstream of SHP2 [38,39,40], and its activation can promote the differentiation of Tregs [41, 42], we further detected the level of p-SMAD2/3. The results showed (Fig. 2H) that the p-SMAD2/3 level was the highest in sFgl2-MSCs, and decreased when CD32b was blocked or in the presence of MSCs or rFgl2 alone.

In conclusion, compared with MSCs or sFgl2 alone, sFgl2-MSCs have a more powerful immunosuppressive effect mediated by the promotion of SHP2 and SMAD2/3 phosphorylation, which can more effectively increase the proportion of Tregs and decrease the proportions of Th1 and Th17 cells.

sFgl2-MSCs significantly alleviate experimental AIH in mice

We found that sFgl2-MSCs had a more significant inhibitory effect than MSCs or sFgl2 alone, and the therapeutic effect of sFgl2-MSCs on AIH was verified in vivo. The concanavalin A (Con A)-induced mouse hepatitis model is an approved experimental AIH model [7]. Before the use of sFgl2-MSCs as an in vivo therapy, we first determined whether MSCs could reach the liver using living organ imaging technology. The results showed that although most of the MSCs were trapped in the lungs, some MSCs reached the liver (Fig. 3A). After 24 h, the liver was collected for histopathological analysis, and the results showed that the tissue damage in the untreated group was the most serious, as indicated by a large amount of inflammatory cell infiltration, large areas of liver necrosis and structural destruction. Compared with that in the untreated group, tissue damage in the MSC treatment group was reduced. In addition, inflammatory cells in the sFgl2-MSC group had the least infiltration and a more complete structure (Fig. 3B and C). In addition, the degree of liver damage was quantified by the Knodell score (Fig. 3D). Consistent with pathological findings, liver damage in mice treated with MSCs was significantly less severe than that in the untreated group (Fig. 3D, P < 0.001). Compared with that in the WT-MSC group, liver injury was further reduced in the sFgl2-MSC-treated group (Fig. 3D, P < 0.001).

Fig. 3
figure 3

sFgl2-MSCs significantly alleviate liver injury in AIH mice. A Live organ imaging confirming that MSCs can reach the liver. B H&E staining showing pathological injuries in each group (20 ×). The black arrows show the positive staining areas. C Data display of liver necrosis area in each group, n = 6. D Liver damage was quantified in each group according to the Knodell scoring system, n = 6. E Serum AST levels in each group were determined by ELISA, n = 6. F Serum ALT levels in each group were determined by ELISA, n = 6. The differences between groups were calculated using ANOVA. ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs; ALT, alanine transaminase; AST: aspartate transaminase

Full size image

AST and ALT can be used to evaluate liver function. Compared with those in the untreated group, the serum levels of AST (Fig. 3E, P < 0.001) and ALT (Fig. 3F, P < 0.001) in mice treated with MSCs were significantly decreased. The serum levels of AST (Fig. 3E, P < 0.001) and ALT (Fig. 3F, P < 0.001) were further decreased in mice treated with sFgl2-MSCs.

Then, ELISA was used to assess inflammation in mice. Compared with that in the untreated group, the expression of anti-inflammatory cytokine IL-10 (Fig. 4A, P < 0.001) was significantly increased in the WT-MSC group, while the expression of the proinflammatory cytokines IL-1β (Fig. 4B, P < 0.001), IL-6 (Fig. 4C, P < 0.001) and TNF-α (Fig. 4D, P< 0.001) was decreased. Furthermore, the expression of IL-10 (Fig. 4A, P < 0.001) was further increased in the sFgl2-MSC treatment group compared with the WT-MSC treatment group, and the expression of IL-1β (Fig. 4B, P< 0.001), IL-6 (Fig. 4C, P < 0.001) and TNF-α (Fig. 4D, P< 0.001) was further decreased.

Fig. 4
figure 4

sFgl2-MSCs significantly regulate cytokines in AIH mice. A. ELISA was used to measure the expression level of IL-10 in the serum of each group. B. ELISA of the serum level of IL-1β in each group. C. ELISA of the serum level of IL-6 in the various groups. D. ELISA of the serum levels of TNF-α in the various groups. The differences between groups were calculated using ANOVA. ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs

Full size image

These experimental results show that sFgl2-MSCs have a more significant therapeutic effect on the treatment of experimental AIH than MSCs alone and can more effectively participate in the regulation of cytokines.

sFgl2-MSCs decrease the proportion of CD4+ and CD8+ T cells in mice with experimental AIH

CD4+ T cells and CD8+ T cells are the main T lymphocytes that infiltrate the liver during AIH [24] and mediate autoimmune hepatocyte injury. Therefore, the proportions of these cell types in the liver and spleen in experimental AIH mice were determined in vivo. The results showed that compared with those in the untreated group, the proportions of splenic CD4+ T cells (Fig. 5A and B, P < 0.001) and CD8+ T cells (Fig. 5D and E, P < 0.001) were significantly decreased by MSC treatment. After sFgl2-MSC treatment, the proportions of CD4+ T cells (Fig. 5A and B, P < 0.001) and CD8+ T cells (Fig. 5D and E, P < 0.001) were further decreased. To better examine the local therapeutic effect of sFgl2-MSCs on the liver, local lymphocytes in the liver were obtained for analysis. The results showed that compared with those in the untreated group, the proportions of liver-derived CD4+ T cells (Fig. 5A and C, P < 0.001) and CD8+ T cells (Fig. 5D and F, P < 0.001) were significantly decreased after treatment with MSCs. The proportions of CD4+ T cells (Fig. 5A and C, P < 0.001) and CD8+ T cells (Fig. 5D and F, P < 0.001) were further decreased in the sFgl2-MSC group.

Fig. 5
figure 5

sFgl2-MSCs reduced the proportions of CD4+ T cells and CD8+ T cells in the spleen and liver in experimental AIH mice. A Dot plots showing CD4+ T cells in the spleen and liver. B The proportions of CD4+ T cells in the spleen in the different groups, n = 6. C The proportions of CD4+ T cells in the liver in the different groups, n = 6. D Dot plots showing CD8+ T cells in the spleen and liver. E The proportions of CD8+ T cells in the spleen in the different groups, n = 6. F The proportions of CD8+ T cells in the liver in the different groups, n = 6. *: P < 0.05, **: P < 0.01, ***: P < 0.001. The differences between groups were calculated using ANOVA. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs

Full size image

These results suggest that sFgl2-MSCs could exert their effects throughout the body, reducing the proportions of CD4+ T cells and CD8+ T cells in the spleen and reducing the infiltration of CD4+ T cells and CD8+ T cells into the liver.

sFgl2-MSCs increase the proportion of Tregs and decrease the proportion of Th17 cells

Tregs, Th1 cells, and Th17 cells all play important roles in the development and treatment of AIH. Therefore, the proportions of these cell types in CD4+ T cells were determined in vivo. Compared with that in the untreated group, the proportion of Tregs (Fig. 6A and D, P < 0.001) in the spleen was increased after MSC treatment, while the proportions of Th17 cells (Fig. 6B and E, P < 0.001) and Th1 cells (Fig. 6C and F, P < 0.001) were decreased. Furthermore, the proportion of Tregs (Fig. 6A and D, P < 0.001) was further increased, while the proportions of Th17 cells (Fig. 6B and E, P < 0.001) and Th1 cells (Fig. 6C and F, P < 0.001) were further decreased after sFgl2-MSC treatment.

Fig. 6
figure 6

sFgl2-MSCs increased the proportions of Tregs in the spleen and liver in AIH mice and decreased the proportions of Th17 and Th1 cells. A Dot plots showing Tregs among CD4+ T cells in the spleen and liver. B Dot plots showing Th17 cells among CD4+ T cells in the spleen and liver. C Dot plots showing Th17 cells among CD4+ T cells in the spleen and liver. D The proportions of Tregs in the spleen and liver in the different groups, n = 6. E The proportions of Th17 cells in the spleen and liver in the different groups, n ≥ 5. F The proportions of Th1 cells in the spleen and liver in the different groups, n ≥ 5. The differences between groups were calculated using ANOVA. *: P < 0.05, **: P < 0.01, ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs

Full size image

Then, the infiltration of Tregs, Th17 cells and Th1 cells in the liver was detected. The results showed that, compared with that in the untreated group, the proportion of Tregs (Fig. 6A and D, P < 0.001) in the liver was increased after MSC treatment, while the proportions of Th17 cells (Fig. 6B and E, P < 0.001) and Th1 cells (Fig. 6C and F, P < 0.001) were decreased. Furthermore, the proportion of Tregs (Fig. 6A and D, P < 0.001) was further increased, while the proportions of Th17 cells (Fig. 6B and E, P < 0.001) and Th1 cells (Fig. 6C and F, P < 0.001) were further decreased after sFgl2-MSC treatment.

To more intuitively observe the distribution of Tregs, Th1 cells and Th17 cells in the local liver, we used immunohistochemical staining to detect the specific markers of the above three T-cell subsets, among which T-bet is considered a Th1 cell marker [43, 44], Foxp3 is considered a marker of Tregs [45, 46], and RORγ is considered a marker of Th17 cells [46, 47]. The results showed (Fig. 7A) that compared with the untreated group, the number of Foxp3-positive cells in MSC-treated livers increased (Fig. 7B, P < 0.001), while the number of T-bet-positive cells and RORγ-positive cells decreased (Fig. 7C and D, P < 0.001). After sFgl2-MSC treatment, the number of Foxp3-positive cells further increased (Fig. 7B, P < 0.001), while the number of T-bet-positive cells and RORγ-positive cells further decreased (Fig. 7C and D, P < 0.001).

Fig. 7
figure 7

sFgl2-MSCs increased liver-infiltrating Foxp3-positive cells and reduced RORγ-positive cells and T-bet-positive cells. A Representative infiltration of Foxp3-, RORγ- and T-bet-positive cells in each group. B Data display of Foxp3-positive cells in each group, n = 6. C Data display of RORγ-positive cells in each group, n = 6. D Data display of T-bet-positive cells in each group, n = 6. The black arrows show the positive staining cells. The differences between groups were calculated using ANOVA. ***: P < 0.001. Abbreviations: MSCs, mesenchymal stem cells; sFgl2, soluble fibronectin-like protein 2; WT-MSCs, wild-type MSCs; NC-MSCs, negative control MSCs

Full size image

These results suggest that sFgl2-MSCs can increase the proportion of Tregs and decrease the proportions of Th17 cells and Th1 cells to reduce the systemic and local immune response to injury.

Discussion

Glucocorticoids alone or in combination with azathioprine are the standard treatment for AIH and are effective in most patients [6]. The treatment end point of AIH is the normalization of serum transaminase and immunoglobulin levels and is used to define complete response [48, 49]. Additional treatments need to be developed for patients who do not have remission or cannot tolerate standard treatments, and stem cell therapy is a potential important treatment. MSCs can regulate the differentiation and activation of almost all immune cells, such as T cells, macrophages and DCs, and are easy to culture, expand and preserve in vitro. Due to their excellent immunomodulatory effects, MSCs have been widely used in the treatment of T-cell-mediated hepatitis [9, 20, 50,51,52]. However, the curative effect of MSCs alone is unsatisfactory, so it is necessary to find ways to enhance the immune regulatory function of MSCs [18,19,20].

FcγRIIb, which is the only inhibitory Fcγ receptor [21], is thought to be present on B cells, macrophages, DCs, and granulocytes. FcγRIIb is associated with immune tolerance and can promote the apoptosis of B cells [53]. In addition, the absence of CD32b increases the risk of autoimmune diseases [54, 55]. Recent studies have shown that this receptor is also present on T cells [22, 56, 57]. In this study, the expression of CD32b on CD4+ T cells increased after murine MSCs were cocultured with CD4+ T cells in vitro. Unlike other Fcγ receptors, the functional ligand of CD32b is sFgl2 [22]. sFgl2 has a wide range of effects on cells: it promotes apoptosis in B cells, inhibits the maturation of DCs, promotes the polarization of alternatively activated macrophages, and is an important molecule that mediates the effect of Tregs [58, 59]. sFgl2 can inhibit the proliferation and differentiation of T lymphocytes, and Fgl2 plays a key role in AIH by regulating Tregs and inhibiting Th17 cells and Th1 cells [60].

Based on the plasticity of MSCs, modifying MSCs to overexpress related factors is a common strategy to enhance the immunomodulatory function of MSCs [61, 62]. Therefore, in this study, the Fgl2 gene was transfected into MSCs by lentiviral transfection technology to induce high sFgl2 overexpression, and the expression of CD32b was induced to the same extent after MSCs were cocultured with CD4+ T cells. Furthermore, we also found that the proportion of Tregs among CD4+ T cells increased after the cells were cocultured with transfected or untransfected MSCs, and the increase was most significant in the sFgl2-MSC treatment group, while the decrease in Th1 and Th17 cells was the most significant. This result suggests that MSCs or sFgl2 alone can exert immunosuppressive effects to different degrees, thereby increasing the proportion of Tregs and decreasing the proportions of Th17 and Th1 cells. In addition, sFgl2-MSCs have a more significant immunoregulatory effect than unmodified MSCs.

Next, we proposed a possible mechanism by which sFgl2-MSCs can more effectively regulate CD4+ T-cell differentiation by regulating the levels of p-SHP2 and p-SMAD2/3. CD32b contains the immunoreceptor tyrosine-based inhibitory motif domain, which is an inhibitory domain [63], and can activate the activation of downstream SHP2 [53]. In fact, Hou et al. [64] reported that sFgl2 activates the ERK1/2-STAT3 pathway by activating SHP2, thereby biasing macrophages toward a pro-repair phenotype, which is associated with endometriosis progression, and our study also revealed that the level of p-SHP2 increased in the presence of sFgl2, and the level of p-SHP2 was the highest in the sFgl2-MSC treatment group because MSCs induced the expression of CD32b in CD4+ T cells; the level of p-SHP2 decreased after blocking CD32b. SMAD2/3 is closely related to the differentiation of Tregs [41, 42] and is also a reported important downstream target of SHP2. This study revealed that MSCs or rFgl2 can increase the level of p-SMAD2/3, which was further increased after sFgl2-MSC treatment. Therefore, sFgl2-MSCs exhibit more powerful immune regulatory function by effectively activating SHP2 and SMAD2/3.

Previous studies have confirmed that MSCs can accumulate in large numbers in the lungs, and some MSCs can migrate to other sites, such as the liver [65,66,67], and MSCs will preferentially gather at the injury site [68, 69]. Our live organ imaging results suggest that although many MSCs accumulate in the lungs, some MSCs reach the liver as the site of injury and play a role.

The efficacy of MSCs accumulated in the lungs is also worthy of attention. Zhang X et al. found that prostaglandin E2 produced by lung MSCs promoted the synthesis and release of acetylcholine, activated cholinergic anti-inflammatory pathways, reduced lung inflammation, and relieved respiratory symptoms in patients with acute respiratory distress syndrome [70]. Furthermore, S.H.M. Pang et al. It was found that apoptosis and exocytosis of lung MSCs can cause changes in alveolar macrophage metabolism and inflammatory pathways, thereby affecting immunosuppression and reducing disease severity [71]. In addition, MSCs have a paracrine function and can secrete a variety of molecules including prostaglandin E2 to play a role [72]. The sFgl2-MSCs transformed in this study can also secrete sFgl2, and after accumulating in the lungs, they can also secrete sFgl2 to play an immunosuppressive function.

In fact, the results of this study showed that compared with the MSCs-treated group, sFgl2-MSCs more significantly protected the liver structure, significantly reduced inflammatory cell infiltration, lowered serum AST and ALT levels, and effectively reduced the pro-inflammatory factor IL-6 compared with the MSCs-treated group., IL-1β and TNF-α levels, while increasing the level of the anti-inflammatory factor IL-10, reducing systemic inflammation, these results suggest that sFgl2-MSCs have a more potent therapeutic effect on experimental AIH.

CD4+ T cells and CD8+ T cells are the main T lymphocytes that infiltrate the liver [1]. The in vivo results confirmed that in experimental AIH mice, CD4+ T cells and CD8+ T cells were significantly elevated in the spleen and liver, indicating systemic inflammation. After sFgl2-MSC treatment in experimental AIH mice, CD4+ T cells and CD8+ T cells were significantly decreased compared with those in the MSC group.

Tregs, which make up 5–10% of all peripheral T cells in healthy humans, control innate and acquired immune responses by limiting tolerance mechanisms and thereby limiting the ability of circulating autoreactive T cells to damage tissue [6]. However, in AIH, the number and functionality of Tregs are decreased [73, 74]. Adoptive transfer of Tregs has shown initial success in treating experimental AIH [75]. Therefore, increasing the number and function of Tregs is critical for treatment [6]. We found that the proportion of Tregs was significantly increased in the spleen and liver after treatment with sFgl2-MSCs compared with MSCs. In addition, increases in the proportions of Th1 cells and Th17 cells, as well as the levels of corresponding cytokines interferon-γ and IL-17A, are characteristic of AIH [4, 76]. The loss of Th17 cells has been shown to reduce T-cell-mediated hepatitis [77]. In this study, we found that compared with MSC treatment, sFgl2-MSC treatment further reduced the proportions of Th17 cells and Th1 cells in the spleen and liver. Therefore, it is reasonable to conclude that sFgl2-MSCs play a significant role in the treatment of AIH by regulating the differentiation of Tregs, Th17 cells and Th1 cells in vivo.

Although this study showed that MSCs increased CD32b expression on CD4+ T cells, the specific mechanism still needs to be further examined. To further improve the therapeutic effect of sFgl2-MSCs, we will further optimize their application strategy in follow-up studies to enhance the curative effect of sFgl2-MSCs on AIH.

Conclusion

sFgl2 gene-modified MSCs may promote the differentiation of Treg cells by enhancing the phosphorylation of SHP2 and SMAD2/3, inhibit the differentiation of Th17 and Th1 cells, and have a significant immune regulation effect (Fig. 8), so as to better treat experimental AIH mice, and it provides a possible strategy for the clinical treatment of AIH.

Fig. 8
figure 8

Possible mechanism of sFgl2 gene-modified MSCs in the treatment of experimental AIH mice. sFgl2 gene-modified MSCs may promote the differentiation of CD4+ T cells into Tregs and inhibit their differentiation into Th17 and Th1 cells by activating SHP2 and SMAD2/3 to treat experimental AIH mice more effectively. AIH: autoimmune hepatitis; MSCs: mesenchymal stem cells; sFgl2: soluble fibronectin-like protein 2; sFgl2-MSCs: sFgl2 gene-modified MSCs; Tregs: T regulatory cells. This image was drawn by the author using Adobe Illustrator 2023

Full size image