MSCs improve gross and microscopic liver histopathology and prolong survival of mice with FHF
Mice were sacrificed 24, 48, and 72 h after MSCs intravenous infusion (Additional file 1: Figure S2A). Nine control and one MSC-treated mouse died before sacrifice. Six of the nine remaining control livers were soft and shrunken with extensive extravasated blood, which was gradually reduced at later time points. By contrast, none of the remaining six MSC–treated livers showed gross pathological changes, even 24 h after MSC infusion (Fig. 1a).
Hematoxylin and eosin (HE)-stained liver sections from control mice revealed dramatical hepatocellular death with cytoplasmic vacuolization, panlobular mononuclear CD45-positive leukocyte infiltration (particularly F4/80-positive macrophages), and severe distortion of liver tissue architecture. By contrast, liver sections from MSC–treated mice rarely showed periportal immune cell infiltration with edema and fibrin deposition (Fig. 1d–f; Additional file 1: Figure S2C–E). Also, flow cytometry analysis of total lymphocytes showed that macrophages infiltrated into control livers more than MSC-treated livers (Fig. 1c; Additional file 1: Figure S2B).
During the 7-day follow-up period after cell transplantation, nine of the 18 control mice successively died, with mortality rate reaching 50 %. By contrast, better survival was observed for MSC–treated mice, with only one mouse dying during the observation period (Fig. 1b).
Overall, these results demonstrate that MSC inhibits the development of histopathological changes and immune cell infiltration and reduces mortality among mice with TAA-induced FHF.
MSC therapy suppresses CCl4-induced chronic liver fibrosis and down-regulates infiltrating macrophages
In order to observe changes in liver fibrosis after mice were treated with CCl4 and MSC, liver sections (Additional file 1: Figure S3) were stained with Sirius Red to identify collagen deposition. Six mice from normal, control and MSC groups were sacrificed for various microscopic evaluations 3 weeks after MSC infusion (Additional file 1: Figure S4A). Sirius Red-stained liver sections revealed massive collagen deposition in livers from control mice (Fig. 2a, f). MSC treatment also noticeably decreased collagen deposition in mice with TAA-induced chronic liver fibrosis, although this effect was less significant than that observed in mice with CCl4-induced chronic liver fibrosis (Additional file 1: Figure S3). The results of immunofluoresence staining of Collagen1 (Col-1) and Collagen3 (Col-3), which are primary contributors to collagen deposition, was consistent with that of Sirius Red staining (Additional file 1: Figure S4B, C, D, E). Also, periodic acid-Schiff (PAS)-staining of liver sections revealed that MSC treatment improved glycogen synthesis and storage (Fig. 2c, h). Moreover, control liver sections around sinus hepaticus were significantly positive for α-smooth muscle actin (α-SMA), a marker of activated HSCs, whereas MSC-treated liver sections showed a sizeable reduction in α-SMA positivity (Fig. 2d, i).
Microscopic evaluation of HE-stained liver sections revealed massive inflammatory infiltration, particularly F4/80-positive macrophages, in livers from control mice. By contrast, MSC treatment markedly down-regulated F4/80-positive macrophage infiltration (Fig. 2b, e, g, j). Taken together, these results demonstrate that MSC therapy suppresses liver fibrosis by down-regulating macrophage infiltration and promoting HSC apoptosis or decreasing the activated HSCs.
MSCs inhibit hepatocellular apoptosis and enhance liver regeneration in vivo
To determine whether MSCs treatment reduces hepatocellular apoptosis, we examined TUNEL-reactive hepatocyte nuclei in liver sections. In control mice with FHF and chronic liver fibrosis, many large apoptotic hepatocyte nuclei were observed, yet only few such nuclei were observed after MSC treatment. Furthermore, the extravasated blood observed after TAA stimulation disappeared after MSC infusion (Fig. 3a, b). Quantification of these observations confirmed a dramatic reduction in TUNEL-positivity in MSC-treated mice with FHF and chronic liver fibrosis (0.10 ± 0.06 or 0.40 ± 0.05 % per field of view, respectively) compared with control mice (1.92 ± 0.40 or 2.83 ± 0.30 % per field of view) (Fig. 3e, f), demonstrating that MSC effectively inhibits hepatocellular death in models of liver failure.
The therapeutic effects of MSCs may rely on the launch of endogenous repair programs. Hepatocytes positive for the proliferation marker Ki67 were quantified in mice with FHF and chronic liver fibrosis and compared with those in control mice (Fig. 3c, d). Whereas few Ki67-positive hepatocytes were observed in control livers (0.08 ± 0.04 or 0.20 ± 0.03 % per field of view), many were observed in MSC–treated liver (2.62 ± 0.50 or 1.62 ± 0.20 % per field of view) (Fig. 3g, h). These findings demonstrate that MSC treatment inhibits hepatocellular apoptosis and stimulates liver regeneration programs in mice with liver failure.
MSC-CM partially ameliorates FHF, but dramatically improves chronic liver fibrosis
MSCs naturally support hematopoiesis by secreting several trophic molecules, including soluble extracellular matrix glycoproteins, chemokines, cytokines, and growth factors. To determine whether MSC-CM plays an important role in improving liver failure, MSC-CM was intravenously infused into mice with FHF or chronic liver fibrosis. Interestingly, the therapeutic effect of MSC-CM infusion was similar to that of MSC infusion, although there were notable differences in their courses of action for FHF. Mice with FHF were sacrificed 24, 48, or 72 h after MSC-CM infusion (Fig. 4a). Six of the eight survival control livers were soft and shrunken with extensive extravasated blood. Microscopic evaluation of HE-stained control liver sections consistently revealed massive hepatocellular death with cytoplasmic vacuolization, hemorrhage and inflammatory infiltration. These severe pathological changes were also observed in MSC-CM-treated livers 24 and 48 h after infusion, although a therapeutic effect was observed 72 h after MSC-CM infusion (Fig. 4d, j, k). However, MSC-CM treatment did not significantly improve survival rate (55.6 % for MSC-CM-treated vs. 44.5 % for control group) (Fig. 4c). Therefore, it is likely that MSC-CM enhances the liver repair system only at later stages of self-recovery.
To access recovery from chronic liver fibrosis, six of normal, control and MSC-CM groups were sacrificed for various microscopic evaluations 3 weeks after MSC-CM infusion (Fig. 4b). Sirius Red-staining, HE-staining and α-SMA immunofluoresence-staining of liver sections revealed that MSC-CM treatment suppressed collagen fiber deposition, inhibited inflammatory infiltration, and promoted activated HSC apoptosis (Fig. 4e–g, l–n). Also, MSC-CM-treated livers exhibited less TUNEL-positivity and more Ki67-positive hepatocyte nuclei, indicating that MSC-CM trophic molecules inhibit hepatocyte apoptosis and enhance liver regeneration (Fig. 4h, i, o, p). Overall, these results show that MSC-CM improves chronic liver fibrosis, but only partially improves FHF.
MSCs promote macrophage line RAW264.7 apoptosis and MSC-CM promotes apoptosis and inhibits proliferation of HSC line LX-2
Macrophages and HSCs have important role in the pathogenesis of FHF and chronic liver fibrosis, respectively. To determine whether MSCs down-regulate macrophages and HSCs apoptosis in vivo via a direct effect of MSCs themselves or an indirect effect of MSC-secreted soluble factors, we examined the effect of MSCs or MSC-CM on in vitro apoptosis and proliferation of RAW264.7 or LX-2. After co-culture of RAW264.7 and MSCs for 48 h, a two-fold increase in apoptosis was observed (16.6 ± 2.0 % co-culture group vs. 7.9 ± 1.2 % control group) (Fig. 5a, d), whereas MSCs and MSC-CM did not effect on the proliferation of RAW264.7, MSC-CM did not promote RAW264.7 apoptosis.
After LX-2 was supplemented with 2 % MSC-CM in co-culture for 48 h, we observed a massive apoptosis of LX-2 (34.5 ± 5.0 in 2 % MSC-CM-treated group vs. 8.3 ± 1.5 % control group). With 8 % MSC-CM, no significant increase in LX-2 apoptosis was observed (Fig. 5c, e). Supplementation with 2 % MSC-CM also suppressed the proliferation of LX-2 (1.88 ± 0.02 in 2 % MSC-CM group vs. 2.80 ± 0.06 % control group) (Fig. 5b, f). Taken together, these results suggest that MSCs themselves directly facilitate macrophage apoptosis, whereas HSC apoptosis and inhibition of proliferation occur via MSC-CM.
MSCs and MSC-CM have anti-inflammatory effects on TAA- and CCl4-stimulated splenocytes, respectively
We examined the subset distribution of CD4+ T lymphocytes in spleens from control, MSC-treated, and MSC-CM-treated mice using flow cytometry. Based on our prior experience, we expected that the subset distribution of CD4+ T lymphocytes in spleens would not be altered by MSC-CM treatment of TAA-stimulate mice. However, MSC infusion down-regulated pro-inflammatory Type 1 T helper (Th1) and Th17 cells (Fig. 6c, d; Additional file 1: Figure S5A, B) and up-regulated anti-inflammatory regulatory T (Treg) cells in mice with FHF (Fig. 6f; Additional file 1: Figure S5D), whereas the distribution of anti-inflammatory Th2 cells- was not significantly changed (Fig. 6e; Additional file 1: Figure S5C). Consistently, the size of spleen in mice from MSC-CM treatment group was smaller than control group mice, which indicates that mice from MSC-CM treatment group were under lower inflammatory state compared with control group (Fig. 6a, b). Therefore, MSCs directly exert immunosuppressive effects in mice with TAA-induced FHF.
By contrast, both MSC and MSC-CM treatment exerted immunosuppressive effects in CCl4-induced chronic liver fibrosis, although better therapeutic effects were observed after MSC-CM delivery. MSC-CM increased levels of Th2 and Treg cells (Fig. 7c, d, h, i), and reduced levels of Th17 cells (Fig. 7b, g), whereas levels of Th1 cells were unchanged (Fig. 6o, j). Moreover, the size of spleen from MSC-CM treatment mice was smaller than control mice (Fig. 7k, l). Comparatively, MSC treatment did not affect Th17 and Treg cells and only slightly alters inflammatory state in mice with chronic liver fibrosis. Also, MSC and MSC-CM treatment substantially down-regulated macrophages in the spleen of mice with acute and chronic liver failure (Figs. 6g, 7e, j; Additional file 1: Figure S5E), consistent with effects observed in the liver. Therefore, in mice with CCl4-induced chronic liver fibrosis, immunosuppressive effects are mainly attributed to MSC-CM. There results demonstrate that MSCs themselves exert immunosuppressive effects in mice with TAA-induced FHF, whereas MSC-CM underlies the immunosuppressive effects in mice with CCl4-induced chronic liver fibrosis.