Mesenchymal stromal stem cells (MSCs) are an attractive therapeutic model for regenerative medicine due to their pluripotency. MSCs are used as a treatment for several inflammatory diseases, including hepatitis. However, the detailed immunopathological impact of MSC treatment on liver disease, particularly for adipose tissue derived stromal stem cells (ADSCs), has not been described. Here, we investigated the immuno-modulatory effect of ADSCs on hepatitis using an acute ConA C57BL/6 murine hepatitis model. i.v. administration of ADSCs simultaneously or 3 h post injection prevented and treated ConA-induced hepatitis. Immunohistochemical analysis revealed higher numbers of CD11b+, Gr-1+, and F4/80+ cells in the liver of ConA-induced hepatitis mice was ameliorated after the administration of ADSCs. Hepatic expression of genes affected by ADSC administration indicated tissue regeneration-related biological processes, affecting myeloid-lineage immune-mediating Gr-1+ and CD11b+ cells. Pathway analysis of the genes expressed in ADSC-treated hepatic inflammatory cells revealed the possible involvement of T cells and macrophages. TNF-α and IFN-γ expression was downregulated in hepatic CD4+ T cells isolated from hepatitis livers co-cultured with ADSCs. Thus, the immunosuppressive effect of ADSCs in a C57BL/6 murine ConA hepatitis model was dependent primarily on the suppression of myeloid-lineage cells and, in part, of CD4+ T cells.
Mesenchymal stromal stem cells (MSCs) are somatic cells that reside in the mesenchymal tissues, such as the BM, umbilical cord, and adipose tissue [1, 2]. MSCs are able to differentiate into several types of cells (pluripotent) in the same lineage, such as chondrocytes, osteocytes, adipocytes, and cardiomyocytes, as well as those of different lineages, such as hepatocytes. Because of this differentiation capability, they have been studied as a possible application in regenerative therapy of miscellaneous impaired organs, such as breast reconstruction  and repair of ischemic heart tissue . Another intriguing characteristic of MSCs is their immunomodulatory potency . Because most liver diseases, including viral hepatitis [6, 7], primary biliary cirrhosis , autoimmune hepatitis , and steatohepatitis , are associated with hepatic inflammatory cells , elucidation of the effect of MSCs on hepatic inflammation is important when considering the use of MSCs for treating liver diseases. Although the efficacy of MSC treatment of liver diseases has been reported , the detailed immunopathological impact of MSC treatment on liver diseases, particularly for adipose tissue derived stromal stem cells (ADSCs), has not been investigated.
ConA, a plant lectin , is frequently used to induce acute hepatitis in rodents  to model the pathological features of autoimmune hepatitis. Although this model is mediated mainly by lymphocyte-lineage cells such as T cells and NKT cells, Kupffer cells/macrophages also participate in hepatitis. Therefore, evaluating the therapeutic efficacy of ADSCs in this murine hepatitis model is important. Although the potential efficacy of ADSCs in a BALB/c ConA hepatitis model has been reported , the immunopathology has not been investigated.
We confirmed that immediate i.v. administration of ADSCs after ConA injection prevented hepatitis. We also observed that administering ADSCs 3 h after the ConA injection resulted in successful treatment of hepatitis, as the liver was already infiltrated by CD11b+ and Gr-1+ inflammatory cells. Gene expression analysis of the liver showed that ADSC treatment affected myeloid-lineage cells, providing repair and regenerative effects in ConA-induced hepatitis mice. Moreover, gene expression analysis of hepatic inflammatory cells indicated pathways related to T cells and monocyte-lineage cells. Pathologically important cytokines such as TNF-α and IFN-γ were upregulated in CD4+ T cells isolated from ConA-induced hepatitis mice but were significantly suppressed by co-culture with ADSCs. Thus, the anti-inflammatory effects of ADSCs in the C57BL/6 murine ConA hepatitis model were mediated by the suppression of myeloid-lineage and CD4+ T cells.
Characteristics of the immune response in ConA-induced hepatitis mice
To examine the characteristics of ConA-induced acute hepatitis, we injected 300 μg ConA into C57BL/6 female mice (n = 4) and determined serum alanine transferase (ALT) and lactate dehydrogenase (LDH) activities. Serum ALT and LDH activities were elevated through 24 h (Fig. 1A). The macroscopic appearance and histology of the liver obtained 24 h after ConA injection revealed intense necrosis (Fig. 1B). The immunohistochemical analysis showed that the number of CD4+ T cells in the liver peaked at 6 h after the ConA injection, and remained high for 24 h (Fig. 1C and D). The numbers of CD11b+ cells and Gr-1+ cells accumulated in the liver increased at 3 h and reached a maximum at 12 h after ConA injection (Fig. 1C and D). The numbers of F4/80+ monocyte/macrophage lineage cells increased at 6 h after the ConA injection, but returned to basal levels after 24 h (Fig. 1C and D). We also assessed the frequency of CD11b+/Gr-1+ cells, as a phenotype of myeloid-derived suppressor cells (MDSCs), in ConA hepatitis mice at 6 h (n = 3). The frequency of CD11b+/Gr-1+ cells was higher than that in WT mice (Fig. 1E). Scavenger receptor CD204 expression was higher in CD11b+/Gr-1+ cells than CD11b+/Gr-1− cells (Fig. 1F), and the population gated for CD11b+ cells contained granulocytic Ly-6C+/Ly-6G+ cells as well as monocytic Ly-6C+/Ly-6G− cells (Fig. 1G).
To determine the type of immune-mediating cells involved in ConA-induced acute hepatitis, we depleted mice of various immune cell subpopulations (n = 4 per group). Mice that were pretreated with clodronate, a reagent that depletes monocyte-macrophage lineage cells , followed by injection of ConA, did not show a significant elevation in serum ALT or LDH activity (Fig. 1H). Mild elevation of serum activity for these enzymes in mice depleted of CD4+ T cells was observed, whereas depletion of CD8+ T cells had no significant effect. These results suggest the importance of monocyte-macrophage myeloid-lineage cells, as well as the contribution of CD4+ T cells, in ConA-induced hepatitis.
ConA-induced acute hepatitis is ameliorated by i.v. administration of ADSCs in vivo
Next, we determined the therapeutic efficacy of ADSCs in the ConA-induced hepatitis model. We obtained and expanded stromal cells from adipose tissue by passaging them eight to ten times (Fig. 2A). Almost all cells expressed the mesenchymal lineage markers, CD29 and CD44 (Fig. 2B). With regard to stem cell markers , approximately 40% and 73% of cells expressed CD105 and CD90, respectively (Fig. 2B). Moreover, the cells were pluripotent and were able to differentiate into osteocytes, chondrocytes, and adipocytes (Fig. 2C–F). When 1 × 105 ADSCs were administered via the tail vein immediately after ConA injection in mice (n = 3), the elevation of serum ALT and LDH activity was substantially ameliorated, compared with mice without ADSC treatment (n = 4) 24 h after injection (Fig. 3A). In terms of therapeutic efficacy, 1 × 105 ADSCs were administered to mice 3 h after ConA injection (n = 3), serum ALT and LDH activities were significantly reduced in acute hepatitis mice treated with ADSCs, compared with ConA-induced hepatitis mice without treatment (n = 4), 24 h after ConA administration (Fig. 3B). The macroscopic
appearance of the liver obtained from mice injected with 300 μg of ConA followed by ADSC administration showed a mild and spotty white area with an almost normal color (Fig. 3C). Liver histology showed an almost normal appearance, with no necrosis (Fig. 3C), indicating that ConA-induced hepatitis was markedly ameliorated by ADSC treatment. No preventive or therapeutic effect on ConA-induced hepatitis resulted from administration of primary cultured murine hepatocytes (n = 3); there was no significant reduction in serum ALT or LDH (Fig. 3A and B), macroscopic necrosis appearance, or histological necrosis, compared with ConA-induced hepatitis (Fig. 3C).
ADSC treatment reduces elevated cytokine/chemokine concentrations in ConA-induced hepatitis mice
Marked protective and therapeutic effects of ADSCs on ConA-induced hepatitis were observed. To determine the effect of ADSC treatment on systemic inflammation in ConA-induced hepatitis, we measured serum cytokine and chemokine concentrations in ConA-induced hepatitis mice treated with ADSCs. Mice injected with ConA were immediately treated with ADSCs and serum was collected 6 h after ConA injection (n = 3). The elevated serum IFN-γ, IL-2, IL-6, IL-4, IP-10, MIG, KC, and MCP-1 levels in ConA-injected mice (n = 3) were significantly reduced by ADSC treatment (Supporting Information Fig. 1A). Injection of mice with ADSCs 3 h after ConA administration (n = 4) resulted in significantly reduced serum IFN-γ, IL-2, IL-6, and MIG levels, compared to ConA-injected mice not treated with ADSCs (n = 6) (Supporting Information Fig. 1B). Thus, the levels of the array of cytokines and chemokines that are elevated in the sera of ConA-induced hepatitis mice were significantly decreased by ADSC treatment.
Distribution of i.v. administered ADSCs in ConA-induced hepatitis murine models
The distribution of administered ADSCs in ConA-induced hepatitis mice was determined by immunohistochemistry. Administered GFP-expressing ADSCs were observed in the lung, but not the liver, of mice injected with ConA followed by immediate ADSC administration (n = 6), through 24 h (Supporting Information Fig. 2A). When administered to mice 3 h after ConA injection (n = 6), GFP-expressing ADSCs were observed primarily in the lung, and a few in the liver (Supporting Information Fig. 2B), suggesting that some fraction of ADSCs reached the liver upon occurrence of hepatitis.
Hepatic gene expression changes by ADSCs treatment are associated with Gr-1+ and CD11b+ cells
To investigate the detailed biological features of the liver in ConA-induced hepatitis mice that were treated with ADSCs, we examined the gene expression profiles of liver tissue of ConA-injected mice obtained 2 h after treatment with ADSCs using a DNA microarray. In the liver tissues of mice treated with ADSCs immediately after ConA injection (n = 3), 589 gene probes were differentially expressed compared with that in mice with ConA-induced hepatitis that had not been treated with ADSCs (n = 3). Expression of the majority of genes was downregulated by ADSCs, as shown by green color (p < 0.05; Fig. 4A). Principal component analysis using these genes showed a discernible distribution difference between the ADSC-treated and -untreated groups (Fig. 4B). When mice received ADSC treatment 3 h after ConA injection, hepatic expression of 309 gene probes was altered significantly compared with those in mice with ConA-induced hepatitis that had not been treated with ADSCs (n = 3). Expression of the majority of genes was downregulated by ADSCs, as shown by green color (p < 0.01; Fig. 4C). Principal component analysis of these genes also showed a discernible distribution difference between the ADSC-treated and untreated groups (Fig. 4D). In the context of biological maps of the genes affected by immediate ADSC treatment, cell differentiation, the inflammatory response, the DNA damage response, and apoptosis predominated (Supporting Information Table 1). In addition to these maps, tissue remodeling and wound repair, mitogenic signaling, and vascular development (angiogenesis) predominated in mice that had received ADSC treatment 3 h after ConA injection (Table 1), indicating that ADSCs provided not only anti-inflammatory effects, but also remodeling effects, in the ConA-damaged liver.
Table 1. Maps relevant to genes for which the expression was affected in the liver of ConA-injected mice followed by ADSC treatment at 3 h
Tissue remodeling and wound repair
Vascular development (angiogenesis)
DNA damage response
Cystic fibrosis disease
Immune system response
Next, we investigated the relevance of these altered genes in the context of inflammatory cells using the public gene expression database of hematopoietic cells and stem cells (GSE27787). The annotated genes among the 589 gene probes detected by microarray analysis probes in the livers of mice that received ADSC treatment immediately after ConA injection were not relevant to any hematopoietic cell type (Fig. 4E). By contrast, among the 309 gene probes, the majority of the annotated genes whose hepatic expression in mice that received ADSC treatment 3 h after ConA injection was affected significantly were found to be highly expressed in Gr-1+ cells and Mac1+ (CD11b+) cells — as indicated by the red color (Fig. 4F). Since majority of the 309 gene probes in the liver of ConA hepatitis were downregulated by ADSC treatment, as indicated by green color (Fig. 4C), these results suggested that effects on Gr-1+ and CD11b+ cells were associated with the therapeutic effect of ADSCs 3 h after ConA injection.
ADSC treatment represses inflammatory cell accumulation in ConA-induced hepatitis
To determine the influence of ADSC treatment on the infiltration/accumulation of immune-mediating cells in the liver of ConA-induced hepatitis mice, we assessed by immunohistochemistry the inflammatory cells in the liver tissues of mice injected with ConA followed by ADSC administration at 3 h. Liver tissues obtained at 6, 12, and 24 h (n = 4 each time point) after ConA injection showed reduced accumulation of CD11b+ cells, Gr-1+ cells, and F4/80+ cells after ADSC treatment (Fig. 5). In contrast, the increased number of infiltrated CD4+ T cells in ConA-induced hepatitis mice was not significantly affected by the ADSCs (Fig. 5). Thus, the predominant change in ConA-induced hepatitis mice treated with ADSCs was in the number of myeloid-lineage inflammatory cells, consistent with the hepatic gene expression data.
T-cell involvement in the altered gene expression of hepatic inflammatory cells by ADSCs treatment
To further assess the anti-inflammatory effects of ADSCs in mice with ConA-induced hepatitis, we isolated hepatic inflammatory cells from mice 2 h after ADSC treatment, which was administered 3 h after ConA injection (n = 2) and from mice not treated with ADSCs (n = 2). A total of 939 genes were differentially expressed in hepatic inflammatory cells from ConA-induced hepatitis mice treated with ADSCs. The gene expression profiles associated with ADSC treatment and ConA-induced hepatitis without ADSC treatment were readily distinguishable (Supporting Information Fig. 3A). Pathway map analysis showed that these genes were relevant to biological pathways of oncostatin M signaling via JAK-Stat or MAPK signaling and CCR5 signaling in macrophages and T lymphocytes in the immune response pathway (Supporting Information Table 2). Network analysis of these genes featured a network consisting of AcRIIA, STAT3, Activin A, FTSJD1, and STAT1 at the top (Supporting Information Table 3), which indicated that pathways involving IL-2 and TNF-α, and the STAT1/STAT3 pathway were also involved (Supporting Information Fig. 3B). These results suggest that T cells, as well as antigen presenting/phagocytosis lineages, were the immune-mediating cell populations affected by ADSC treatment.
ConA-activated CD4+T cells and CD11b+ cells in the liver are important targets of ADSC treatment
The above data indicated that ADSCs administered in ConA-induced hepatitis had therapeutic immunological effects in terms of repairing the damaged liver and affected CD11b+ and Gr-1+ myeloid-lineage cells and T cells. To further explore how ADSCs affected the subpopulations of inflammatory cells involved in ConA-induced hepatitis, we investigated the expression of cytokine/chemokine-related genes in CD4+ T cells and CD11b+ cells obtained from livers with ConA-induced hepatitis (n = 4) that had been treated in vitro with ADSCs (n = 3). Expression of TNF-α, IL-10, and CXCL10 was significantly downregulated by ADSC treatment in both CD4+ T cells (Supporting Information Fig. 4A) and CD11b+ cells (Supporting Information Fig. 4B). IFN-γ, IL-4, and CXCL9 expression by CD4+ T cells were significantly affected by ADSCs. Although CCL3, which was upregulated by ConA injection, was not significantly affected by ADSCs, the expression of its cognate receptor, CCR5, was decreased in CD4+ T cells (Supporting Information Fig. 4A), suggesting an effect on the CCL3-CCR5 axis. These results suggest that CD4+ T cells and myeloid-lineage CD11b+ cells were the susceptible hepatic inflammatory subpopulations of cells in the ConA-induced hepatitis liver.
Anti-inflammatory effect of ADSCs on ConA hepatitis do not rely on MDSCs
We further assessed whether the anti-inflammatory effect of ADSCs in ConA hepatitis relied on MDSCs. Neither the frequency of nor the NO production by CD11b+Gr-1+ cells were increased by ADSC treatment (Supporting Information Fig. 5A). CD11b+Gr-1+ cells from ConA-injected mice treated with ADSCs showed arginase activity similar to that in CD11b+Gr-1+ cells from ConA-injected mice (Supporting Information Fig. 5B). CD11b+Gr-1+ cells from ConA-injected mice treated with ADSCs suppressed the ConA-stimulated proliferation of T cells in vitro, although the effect was slightly attenuated compared to that of cells from mice with ConA-induced hepatitis (Supporting Information Fig. 5C). Thus, ADSC treatment was not dependent on MDSCs induced by ConA hepatitis.
MSCs are effective for immune-mediated disease treatment including the ConA-induced BALB/c murine hepatitis model , but the detailed mechanisms have not been fully elucidated. Here, we confirmed that ADSCs have preventive and therapeutic effects in a ConA-induced C57BL/6 hepatitis murine model and assessed the immunopathological mechanisms by determining the participating hepatic immunomodulatory cells. ADSCs injected via the tail vein were found in the lung; some were observed in the liver but only when ADSCs were administered 3 h after ConA injection, a time at which infiltration of CD11b+ and Gr-1+ inflammatory cells into the liver had already begun. Gene expression analysis of liver tissue from ConA-induced hepatitis mice showed that the ADSC treatment induced biological pathways indicative of liver repair and regeneration. Myeloid-lineage cells were the predominant population in terms of affected genes, consistent with immunohistochemical staining of the liver for immune-mediating cells. Furthermore, the gene expression profiles of hepatic inflammatory cells from ConA-induced hepatitis mice treated with ADSCs suggested T-cell and macrophage involvement. Moreover, the expression patterns of cytokine/chemokine-related genes in hepatic inflammatory cells co-cultured with ADSCs suggested that CD4+ T cells were important in ConA-induced hepatitis and were affected by ADSC treatment.
The immunopathological features of ConA-induced hepatitis have been characterized as being primarily lymphocyte-lineage cell-mediated hepatitis [18-20], leading to massive hepatocellular degeneration, necrosis, and apoptosis ; thus, this model is relevant to clinical autoimmune hepatitis. Additionally, Kupffer cells play an important role in induction of hepatitis . Unexpectedly, we observed prominent increases in CD11b+, Gr-1+, and F4/80+ cells in liver tissues of the ConA-induced hepatitis mice. Additionally, we found that the monocyte-macrophage lineage cells contributed most significantly to hepatitis, as confirmed by depletion treatment, such that hepatitis was almost completely abolished when those cell types were abrogated by clodronate. This is further evidenced by the fact that ADSC treatment reduced the number of CD11b+, Gr-1+, and F4/80+ cells in the liver of ConA-induced hepatitis mice (Fig. 5). The importance of Gr-1+ and CD11b+ cells was also suggested by changes in the gene expression profile of the liver of ConA-induced hepatitis treated with ADSCs (Fig. 4C and F). Thus, monocyte-macrophage lineage cells are important in the pathogenesis of ConA-induced hepatitis in mice and are important targets of ADSCs. CD4+ T cells were also involved since their depletion partially ameliorated ConA-induced hepatitis. The number of infiltrating CD4+ T cells in the liver of ConA-induced hepatitis mice was not markedly reduced by ADSC treatment. However, gene expression analysis of hepatic inflammatory cells in ConA-induced hepatitis mice treated with ADSCs showed that signaling of oncostatin M, a type I cytokine associated with developing T cells , and CCR5 signaling in macrophages and T lymphocytes were affected. Therefore, CD4+ T cells participate as an immune mediator and therapeutic target of ADSCs in the pathology of ConA-induced hepatitis mice.
With regard to cytokine/chemokine-related gene expression in hepatic inflammatory cells of ConA-induced hepatitis mice, expression of TNF-α, IL-10, and CXCL10 in CD4+ T cells and CD11b+ cells was downregulated by ADSC treatment (Supporting Information Fig. 4). Additionally, IFN-γ, IL-4, and CXCL9 were also significantly downregulated in CD4+ T cells, but not in CD11b+ cells (Supporting Information Fig. 4). Changes in the expression of the Th2 cytokines, IL-10 and IL-4, were considered to be the secondary consequence of ConA-induced hepatitis, mediated by TNF-α and/or IFN-γ, which are characterized as Th1-associated cytokines . CCR5 expression by CD4+ T cells was downregulated by ADSCs, which may be relevant to the biological processes indicated by the downregulated genes in hepatic inflammatory cells. Because CCR5 is a CD4+ T-cell receptor that interacts with APCs, such as macrophages , suppression of CCR5 expression on CD4+ T cells by ADSC might explain the amelioration of ConA-mediated hepatitis. Overall, the therapeutic efficacy of ADSCs impacted both CD4+ and CD11b+ cells in terms of alteration of levels of inflammatory humoral mediators and cytokine/chemokine profiles, thus contributing to amelioration of ConA-induced hepatitis.
A proportion of i.v. administered ADSCs were present in the livers of ConA mice injected with ADSCs at a time point at which the liver had already been infiltrated with Gr-1+ and CD11b+ cells, whereas no ADSCs were present in the livers of mice injected with ConA following immediate treatment with ADSCs. This indicates that a liver undergoing inflammation attracts administered ADSCs. The extent of inflammation required to recruit ADSCs should be clarified, as it has previously been reported that hepatitis occurring just 30 min after ConA injection results in recruitment of a substantial number of stem cells to the liver in the BALB/c ConA hepatitis model . Given that the migratory capabilities of MSCs are well known although not yet fully investigated , how ADSCs are recruited to an already inflamed liver as a result of ConA administration should be examined. In addition, the ADSCs administered to C57BL/6 mice immediately after ConA injection resided in the lung. In spite of the fact that they were not detected in the liver, these ADSCs prevented ConA hepatitis, indicating the remote effect of ADSCs. Thus, indirect mediators produced by ADSCs associated with their anti-inflammatory effects should be investigated intensively.
In conclusion, the therapeutic anti-inflammatory efficacy of ADSCs relied on suppression of myeloid-lineage and CD4+ T cells in the ConA-induced C57BL/6 murine hepatitis model. The application of ADSC therapy to various inflammatory liver diseases can be further developed by studies of their immunomodulatory effects.
Materials and methods
Murine acute hepatitis induced by ConA injection and treatment with ADSCs
C57BL/6J female mice (10–12 weeks old, Charles River Laboratories Japan Inc., Yokohama, Japan) were injected i.v. with 300 μg of ConA (Sigma-Aldrich, St. Louis, MO, USA) dissolved in PBS. For CD4+ T-cell or CD8+ T-cell depletion, 200 μg of purified anti-CD4 antibody from the culture supernatant of GK1.5 cells (ATCC, Manassas, VA, USA), or purified anti-CD8 antibody from the culture supernatant of 2.43 cells (ATCC), was injected i.p. for two consecutive days before ConA injection. For depletion of monocyte-macrophage lineage cells, 2 mg of clodronate (Sigma-Aldrich), which was encapsulated in liposomes using the COATSOME-EL-01-N liposome formulation kit (Nihonyushi, Tokyo, Japan) , was injected via the tail vein 2 days before ConA injection. For the prevention or treatment experiment, 1 × 105 ADSCs were administered i.v. immediately or 3 h after ConA injection. In some cohorts, blood was obtained under anesthesia, and liver and lung tissues were collected after euthanizing mice at 6, 12, and 24 h after ConA injection. A portion of the liver tissue was homogenized and the enriched fraction of inflammatory cells was obtained by gradient centrifugation using Ficoll–Hypaque (Sigma-Aldrich). Our institutional review board approved the care and use of laboratory animals in all experiments.
Isolation and culture of ADSCs and primary hepatocytes
Inguinal adipose tissues were obtained from C57BL/6J male mice (10–12 weeks old, Charles River Laboratories Japan Inc.) or GFP-transgenic mice (male, 10–12 weeks old, gift from Prof. Okabe, Osaka University, Japan). Tissues were digested with 0.075% collagenase type I (Wako Pure Chemical Industries Ltd., Osaka, Japan), washed with PBS, and then transferred to a culture dish with DMEM/F-12 1:1 medium (Life Technologies–Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated FBS and 1% antibiotic–antimycotic solution (Life Technologies). Cells were maintained and expanded by eight to ten passages before use.
To obtain primary hepatocytes, C57BL/6J male mice (10–12 weeks old) were anesthetized by i.p. injection of pentobarbital (50 mg/kg; Kyoritsu Seiyaku, Tokyo, Japan) and injected with 10 mL of 0.75% type I collagenase solution via the portal vein. Liver tissues were minced to dissociate cells, filtered through a 100 μm mesh, and cultured in 10-cm culture dishes for 16 h until use.
Pluripotency of ADSCs
The pluripotency of ADSCs was examined using a mouse mesenchymal stem cell functional kitⓇ (R&D Systems, Minneapolis, MN, USA), and immunohistochemical staining of cells that had differentiated into osteocytes, chondrocytes, and adipocytes was performed using anti-mouse ostepontin, anti-mouse collagen II, and anti-mouse FABP4 antibodies, respectively, in accordance with the manufacturer's instruction. Adipocyte differentiation was also assessed by staining using an aliquot of Oil Red O (WAKO).
Co-culture of ConA-stimulated hepatic inflammatory cells with ADSCs
Hepatic inflammatory cells were isolated from C57BL/6J female mice (10 weeks old) that had been injected i.v. with 300 μg of ConA 3 h before (n = 4). CD4+ T cells and CD11b+ cells were separated from the collected hepatic inflammatory cells using anti-CD4 and anti-CD11b magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany). Then, 20000 ADSCs were co-cultured with 4 × 105 of the isolated CD4+ T cells or CD11b+ cells in a 24-well plate (BD Falcon, San Jose, CA, USA) for 2 h (n = 3). After co-culture, floating cells were harvested, and RNA harvested using the MicroRNA isolation kit (Stratagene, La Jolla, CA, USA) for real-time PCR analysis to measure cytokine/chemokine gene expression.
Measurement of serum ALT and LDH activity
Blood was collected from the postorbital venous plexus and serum was separated from clotted blood after coagulation. The serum activity of ALT, and LDH was measured using l-type WAKO GPT J2, and LDH-J kits (Wako Pure Chemical Industries Ltd.), respectively, using autoanalytical equipment (Hitach7180, Hitachi Ltd., Tokyo, Japan), according to the manufacturer's protocol.
Measurement of serum cytokine/chemokine concentrations
Sera were obtained from ADSC-treated mice immediately or 3 h after ConA injection (n = 3 and n = 4, respectively), and from ConA-injected mice not treated with ADSCs (n = 3 and n = 6, respectively) at 6 h. Serum concentrations of cytokines and chemokines were measured using a Multiplex Bead Immunoassays kit, Mouse Cytokine 20-Plex Panel (Invitrogen, Carlsbad, CA, USA), following the manufacturer's protocol. The kit covers FGF-basic, GM-CSF, IFN-γ, IL-1α, IL-1β, IL-2, IL-4, IL-5, IL-6, IL-10, IL-12 (p40/p70), IL-13, IL-17, IP-10(CXCL10), KC, MCP-1, MIG(CXCL9), MIP-1α, TNF-α, and VEGF.
Histological and immunohistochemical analyses of liver and lung tissues
Harvested liver and lung tissues were fixed in 10% formaldehyde, embedded in paraffin, sectioned at 4 μm, and stained with H&E. For immunohistochemical analysis, the liver tissues were embedded in OCT compound (Sakura Finetek, Torrance, CA, USA), snap-frozen in liquid nitrogen, cryostat-sectioned, and fixed with methanol/acetone (1:1). The paraffin-embedded tissues were also sliced into 4 μm sections, mounted on microscope slides, and deparaffinized, followed by epitope retrieval using proteinase K (Dako, Glostrup, Denmark). The slides were incubated with peroxidase blocking reagent (Dako) for 15 min at room temperature to inhibit endogenous peroxidase activity, followed by incubation with protein blocking reagent (Dako) to avoid nonspecific protein reactions. The slides were incubated with primary antibodies (anti-mouse CD4, CD11b, Gr-1, F4/80) (BD Pharmingen, San Diego, CA, USA) and anti-GFP (MBL, Nagoya, Japan) diluted with PBS containing 1% BSA overnight at 4°C. After washing in PBS, the slides were then incubated with secondary antibodies (anti-rat, anti-rabbit; Nichirei, Japan) for 30 min at room temperature. The immune complexes were visualized using EnVision kits /HRP (DAB; Dako) followed by counterstaining with hematoxylin. The numbers of positive cells in each section were counted in four randomly selected fields at 100× magnification under a microscope.
RNA isolation and gene expression analysis by DNA microarray
Total RNA was obtained from the tissues or hepatic inflammatory cells in RNAlater (Ambion) using RNA isolation kit (Sigma-Aldrich) in accordance with the supplied protocol with slight modifications. Isolated RNA was amplified and labeled with the Cy3 using the Quick Amp labeling kit (Agilent Technologies, Santa Clara, CA, USA) in accordance with the manufacturer's protocol. cRNA of 1.65 μg was hybridized onto a Whole Mouse Genome 4 × 44K Array (Agilent Technologies) and scanned using a DNA Microarray Scanner (model G2505B, Agilent Technologies).
Gene expression data were analyzed using the GeneSpring analysis software (Agilent Technologies). Each measurement was divided by the 75th percentile of all measurements in that sample at per chip normalization. Hierarchical clustering and principal component analysis of gene expression was performed. Welch's t-test, with Benjamini and Hochberg's false discovery rate, was used to identify genes that were differentially expressed in the groups of interest. Analysis of biological processes was performed using the MetaCore software suite (GeneGo, Carlsbad, CA, USA). BRB array tools (http://linus.nci.nih.gov/BRB-ArrayTools.html) were also used for unsupervised or one-way clustering analyses. Microarray data were deposited in the NBCI Gene Expression Omnibus (GSE ID: GSE41465).
Cultured ADSCs were incubated in PBS supplemented with 2% BSA (Sigma-Aldrich) containing antibodies labeled with FITC or PE anti-mouse CD44 or CD90 (Beckman Coulter, Brea, CA, USA), and CD105 (Miltenyi Biotec) antibodies. Hepatic inflammatory cells isolated from mice were incubated with a mixture of FITC-labeled anti-mouse CD204 (AbD Serotec, Raleigh, NC, USA), PE-labeled anti-mouse Gr-1 (Miltenyi Biotec), and allophycocyanin-labeled anti-mouse CD11b (BioLegend, San Diego, CA, USA), or FITC-labeled anti-mouse CD11b (Beckman Coulter), PE-labeled anti-mouse Ly-6G (BioLegend), and allophycocyanin labeled anti-mouse Ly-6C (BioLegend) antibodies. The fluorescence intensity of the cells was measured using a FACSCalibur™ (Becton Dickinson, San Jose, CA, USA). Data obtained were visualized and analyzed using the FlowJo software (Tomy Digital Biology Co., Ltd., Tokyo, Japan).
Isolation of CD11b+Gr-1+ hepatic inflammatory cells and T-cell [3H]-thymidine incorporation assay
C57BL/6J female mice were injected with 300 μg of ConA and then injected with 1 × 105 ADSCs after 3 h (n = 3). Three hours later, hepatic inflammatory cells were isolated and incubated with FITC-labeled anti-mouse CD11b (Beckman Coulter) and PE-labeled anti-mouse Gr-1 (Miltenyi Biotec) antibodies. The CD11b+Gr-1+ population was collected using a FACSAira II™ (Becton Dickinson). CD11b+Gr-1+ cells (1 × 105), which had been irradiated with 2000 rads, were co-cultured with 1 × 105 purified splenic T cells isolated from C57BL/6J mice in RPMI1640 medium (Invitrogen) supplemented with 10% heat-inactivated FBS, 1% antibiotic–antimycotic solution (Life Technologies), and ConA (4 μg/mL) for 48 h (n = 4). The culture was pulsed with [3H]thymidine (1 μCi/well) for 16 h and harvested. Thymidine incorporation was measured using a beta-counter (PerkinElmer, Waltham, MA, USA).
C57BL/6J female mice were injected with 300 μg of ConA. Three hours later, 1 × 105 ADSCs were injected via the tail vein. After a further 3 h, hepatic inflammatory cells were isolated from ConA hepatitis mice with or without ADSC treatment (n = 3 each) and incubated in PBS supplemented with 2% BSA, PE-labeled anti-mouse Gr-1 antibody, and allophycocyanin-labeled anti-mouse CD11b antibody. Cells were then incubated in PBS containing 2.5 mg/mL diaminofluorescein-FM diacetate (Sekisui Medical Co., Ltd., Tokyo, Japan), which emits fluorescence at 515 nm in a reaction with NO, at 37°C for 30 min and subjected to FACS analysis using a FACSCalibur flow cytometer.
Female C57BL/6J mice were injected with 300 μg of ConA. Three hours later, 1 × 105 ADSCs were injected via the tail vein. After further 3 h, hepatic inflammatory cells were isolated from ConA hepatitis mice with or without ADSC treatment (n = 3 each) and were lysed with PBS containing 10 mM Tris-HCl (pH 7.4) and 0.4% Triton X-100, supplemented with the proteinase inhibitor cocktail, cOmplete, Mini, EDTA-free® (Roche, Basel, Switzerland). One hundred micrograms of the lysis aliquot obtained were subject to an arginase activity assay using a QuantiChrom™ Arginase Assay kit (BioAssay Systems, Hayward, CA), which measures urea produced from the substrate, in accordance with the manufacturer's protocol.
All data are expressed as means ± SE. Statistical analyses were performed using the JMP software (ver.9.02; SAS Institute Japan Inc., Tokyo, Japan). Student's t-test and Wilcoxon signed-rank test were used. p values < 0.05 were considered to indicate statistical significance.
This study was supported, in part, by subsidies from the Japanese Ministry of Education, Culture, Sports, Science, and Technology and the Japanese Ministry of Health, Labor, and Welfare.
Conflict of interest
The authors declare no financial or commercial conflict of interest.