Regulation Of Immune Cell Differentiation And Function By Liver Stromal Microenvironment And The Underlying Mechanisms

The liver is a unique organ in which induction of tolerance is favoured over induction of immunity. For examples, administration of antigens via the portal vein is apt to induce immune tolerance, allogeneic liver transplantation can be established and maintained even without immunosuppression. Also, some pathogens, including malaria, hepatitis B virus and hepatitis C virus, usually cause chronic infection in liver despite the initiation of the immune response. The phenomenon of "liver tolerance" has drawn much attention, whereas related mechanisms are not fully understood.Up to now, the studies on the mechanisms of liver tolerance mainly focused on lymphocytes and antigen-presenting cells (APCs) in liver. The NK and NKT cells rich in liver can secret chemokines to trap the activated T cells which are induced to undergo cell death in liver, which might be one reason for liver tolerance. Another explanation might be that some DCs in liver secreting IL-10 can induce tolerance. However, such kinds of cells exist all over the body, why and how do they preferentially induce tolerance in liver? The answers to these questions point to the unique roles of the liver microenvironment in controlling immunity or maintaining tolerance, which are largely unknown.As a heterogeneous population of APCs, DCs play pivotal roles in the initiation of immunity and induction of immunological tolerance depending on their maturation state and different subsets. In recent years, DCs with regulatory functions have attracted much attention, which are shown to be able to inhibit T cell response and inflammation. On the basis of long-term culture system of splenic stromal cells by O'Neill's group, we and others showed that splenic microenvironment is important in the differentiation of regulatory DCs because haematopoietic progenitors, as well as mature DCs, can be induced by splenic stroma to differentiate into regulatory DCs. However, whether the microenvironment of other organs such as liver has similar effect on the generation of regulatoryDCs is unknown.Liver stroma can support haematopoiesis in embryonic period and sustain focal intrahepatic extramedullary haematopoiesis in adult life in disease states. Also, some experiments further showed that transferred bone marrow cells could localize in liver and differentiate into mature hepatocytes, kupffer cells and hepatic stellate cells, et al. On the basis of our previous studies and these observations that cells with progenitor properties are present in liver and liver stroma is functional in driving differentiation of many kinds of cells, we wonder what the effect of liver stroma on the differentiation of progenitors is, and whether the differentiated cells can contribute to liver tolerance. Furthermore, there are abundant immune cells in liver, we wonder what the effect of liver stoma on their functions is, and whether the effects will contribute to the induction and maintenance of the liver tolerance?First, we obtained liver stromal cells from neonatal liver of C57BL/6 mice by using an organotypic slice culture and analyzed their characteristics. The results of immunohistochemical analyses showed that such liver stromal cells expressed vimentin and desmin, but not cytokeratin 7, and they did not secrete albumin. Furthermore, LDL phagocytosis assay results also demonstrated that such cells did not have characteristics of macrophage or LSECs. The results of RT-PCR indicated that mRNA of a-SMA, a1 (1) collagen, TLMP-1 and MMP-2 were positive in these cells. These results suggested that such liver stromal cells had fibroblast characteristics. Fibroblasts mainly distribute in the portal area and are the major type cells in liver stroma. Under some occasions, such as infection, inflammation and liver damage, hepatic stellate cells can also be activated and transformed into fibroblasts. So, fibroblastic liver stroma can be used to mimic the liver microenvironment to study the relationship between the liver microenvironment and liver immune tolerance.To find out the effects of the liver microenvironment on the progenitors, we co-cultured liver stromal cells with bone marrow-derived hematopoiesis progenitors, and found that liver stromal cells can induce hematopoiesis progenitor differentiation into a kind of dendritic cell-like cells, which had immune regulatory functions to inhibit the proliferation of mature DC-primed antigen-specific CD4⁺ T cells ( we nominated them as liver regulatory dendritic cells, i.e LRDCs). Such cells had myeloid cell characteristics and stretched out dendrites under the phase-contrast microscope. Thephenotype assay showed that such cells did not express CD4, CD8a, B220, Gr-1, CD14 and B7DC; but had low expression of CD11c, I-A^b, CD54, CD40, CD86; and intermediate or high expression of CD11 b, CD80, B7H-1, similar with the phenotype of immature conventional DC (cDCs). LRDCs had more potent phagocytosis ability than cDCs. As a kind of professional APCs, the most important function of DCs is to present antigen to T cells, and induce T cell activation and proliferation. So, we first studied the antigen-presenting capability of LRDCs by coculture with CD4⁺ T cells from OVA_323-339 antigen specific TCR transgenic mice in the presence of OVA peptide, and found that LRDCs can induce activation but not proliferation of OVA_323-339 peptide-specific CD4⁺ T cells.Toll-like receptors (TLRs) are pattern-recognition receptors that trigger innate immune responses, providing both immediate protection against various pathogens and instructing the adaptive immune system by the induction,recruitment and maturation of DCs. As the ligand of TLR4, LPS can bind TLR4 on cDCs, and induce maturation of cDCs with upregulation of MHC-II molecule expression and IL-12 secretion. Whereas, LPS stimulation could not change the phenotype of LRDCs but promoted the secretion of more IL-10 and less IL-12, which similar with the characteristics of regulatory DC reported previously by us. So, we wonder if LRDCs had similar regulatory function to inhibit proliferation of CD4⁺ T cells. By using CFSE-labelled OVA_323-339 antigen-specific CD4⁺ T cells, we analyzed the function of LRDCs in vitro and in vivo, and the results indicated that LRDCs obviously inhibited cDC-induced proliferation of OVA_323-339 antigen-specific CD4⁺ T cells, suggesting that LRDCs exhibit immune regulatory function.We next analyzed mechanisms by which liver stroma programs differentiation of LRDCs from progenitors, and LRDCs inhibit the proliferation of antigen-specific CD4⁺ T cells primed by cDCs. Firstly, by using transwell co-culture system and purified stromal cell member protein, we found that both stromal cell member protein and culture supernatant had effects on LRDCs differentiation. Moreover, we confirmed that blocking M-CSF in supernatant could reduce efficiency of LRDCs differentiation. So, these results indicated that both liver stromal cell supernatant and cell-cell contact can induce LRDCs differentiation, and stromal cell-derived M-CSF is involved in the differentiation of LRDCs. To further elucidate the mechanisms for the inhibitory effect of LRDCs on CD4⁺ T proliferation, we added the neutralizing antibodies to IL-10, TGF-β, B7H-1 and inhibitors to PGE₂ and IDO into the co-culture system of LRDCs and CD4⁺ T cells primed by mature cDC, andfound that PGE₂ inhibition could significantly reverse the inhibitory effect of LRDCs on mature cDC-activated CD4⁺ T cell proliferation, indicating that PGE₂ is involved in the inhibitory effect of LRDCs on CD4⁺ T cell proliferation.IFN-γ is secreted by Th1 cells and is conventionally thought to be responsible for driving cell-mediated immune responses. Overactivated Th1 responses usually cause autoimmune diseases. By contrast, accumulating evidence suggests that IFN-γ can also have anti-inflammatory and regulatory properties, either by acting as Th2 responses, triggering the production of cytokines that counter the activity of Th1 cells, or inducing apoptosis of the activated T cells and inducing generation of regulatory T cells. In this study, we found that LRDCs inhibited proliferation of OVA_323-339 antigen specific CD4⁺ T cells but accompanying with high IFN-γ production. Therefore, we wondered if the high IFN-γ was associated with LRDCs' inhibitory effect on the proliferation or apoptosis induction of CD4⁺ T cells. So we analyzed CD4⁺ T cell apoptosis by using annexin-V and 7-AAD labeling, and found that LRDCs induced apoptosis of activated CD4⁺ T cells more markedly in cDC/CD4⁺T/LRDC co-culture system. Moreover, LRDCs-mediated inhibition of T cell proliferation and induction of CD4⁺ T cell apoptosis could be partially reversed by neutralizing anti-IFN-γ antibody. This data indicated that induction of high IFN-γ is one of the mechanisms for LRDCs to negatively regulate immune response. To further prove that LRDCs can exhibit negative immune regulatory function in vivo, we prepared mouse model with experimental autoimmune hepatitis (EAH) by immunization with allogeneic mice liver antigen S100, and then observed the effect of adoptively transferred LRDCs on the pathogenesis of such EAH mice by analyzing the serum and pathological indications. We found that pathological lesion of EAH mice could be reduced by transferring LRDCs. So, LRDCs might be one of the effector cells that construct liver immune tolerance and maintain liver immune homeostasis.Considering that the above data are just about biological characteristics of the differentiated LRDCs prepared in vitro, we next investigated if the natural counterpart of LRDCs existed in vivo. We analyzed liver nonparenchymal cells (NPCs) of C57BL/6 mice on the basis of the phenotype of LRDCs, and only about 5.90% of CD11b^hi liver NPCs are CD11 c^lowIA-b^lowcells, which express similar cytokine profile and exhibit inhibitory effect on CD4⁺ T cell proliferation, just like that of LRDCs. By using flow cytometer sorter, we demonstrated that such CD11c^lowIA-b^low cells in CD11b^hi liver NPCs might be the natural counterpart of LRDCs in liver. Given that naturalcounterpart can also be differentiated from hematopoiesis progenitor similar with LRDCs cultured in vitro, we also established in situ differentiation model in liver. By transferring progenitors through the superior mesenteric vein injection, bone marrow-derived hematopoiesis progenitors were localized in liver and differentiated into kinds of cells in situ, including LRDCs' natural counterpart. So, these data have confirmed that there is a kind of natural counterpart of LRDCs in liver, which can also be differentiated from hematopoiesis progenitors in situ.Moreover, the liver is also an organ with its own unique subset of resident lymphocytes, especially having the highest frequency of NK cells of all organs. Simultaneously, as a major organ for the invasion and replication of many viruses, the number of NK cells can be increased after virus infection. We wonder why so many NK cells reside in liver? Two hypotheses can be used to explain why the liver contains much more NK cells than other organs. One is local generation or differentiation of NK cells within liver. The other is the strong recruitment of NK cells from the periphery that gives rise to the intrahepatic pool.As we described above that fibroblasts, one of liver stromal cells, exist widely in the liver and can be activated after virus infection or liver injury. So, we used liver fibroblasts to mimic the liver stromal microenvironment to explore its role in the regulation of NK differentiation and functions. When bone marrow-derived Lin^-CD117⁺ progenitors cocultured with liver fibroblast stroma in the presence of exogenous IL-15, part of progenitors could be differentiated into NK1.1 positive cells, which secreted IFN-γ and expressed both perforin and granzyme B. Moreover, a small fraction of such differentiated NK1.1 cells simultaneously expressed Ly49A. Further experiments showed that the differentiation NK1.1 positive cells from Lin^-CD117⁺ progenitors mainly depended on the soluble factors secreted by liver stroma together with exogenous IL-15. So, according to the characteristics of differentiated cells, we concluded that liver fibroblast stroma, working with exogenous IL-15, can induce Lin^-CD117⁺ progenitor differentiation into immature NK cells.Furthermore, we also used liver fibroblast stroma to mimic liver microenvironment and explored the effects of liver microenvironment on the migration and activation of NK cells in liver, especially under the condition of virus-induced hepatitis. Polyl.C is an artificial mimic of viral RNA, which may trigger the immune response through activation of TLR3 signaling. So, we used Poly I:C to mimic virus infection to activate NK cells, and explored the effects of liver microenvironment on theactivation of NK cells. We found that NK cells expressed TLR3 and could be effectively activated by Poly I:C. Liver stroma could recruit resting NK cells, and recruit Poly l:C-activated NK cells more significantly. Also, liver stroma could further promote activation of NK cells stimulated with Poly I:C, leading to increased IFN-γ secretion and cytotoxicity of NK cells. We demonstrated that fibronectin expressed by liver stroma contributes to the enhancement of NK cell activation by liver stroma.In summary, we obtained liver stromal cells using an organotypic slice culture, and found such stromal cells are fibroblasts and can induce differentiation of bone marrow-derived hematopoiesis progenitor into regulatory DCs. Such kind of liver regulatory DCs (LRDCs) exhibit with low CD11c, MHC II but high CD11 b expression, high IL-10 but low IL-12 secretion.. LRDCs also exhibit more potent phagocytosis ability than cDCs. LRDCs can obviously inhibit cDC-induced proliferation of antigen-specific CD4⁺ T cells both in vitro and in vivo, and also induce apoptosis of activated T cells and alleviate the damage of autoimmune hepatitis. We demonstrate that stromal cell-derived M-CSF is involved in the differentiation of LRDCs. DCs-derived PGE₂ and T cells-derived IFN-γ are responsible for the regulatory function of LRDCs. Importantly, natural counterpart of such regulatory DCs is identified in liver and progenitors can be differentiated into such DCs in liver once transferred into liver. Therefore, liver microenvironment can program progenitors to differentiate into regulatory DCs in situ, which contributes to liver tolerance. In another part of our study, we also demonstrate that such liver stromal cells can program progenitors to differentiate into immature NK cells in the presence of exogenous IL-15, and chemoattract NK cells and further enhance activation of NK cells stimulated with Poly I:C, indicating that liver microenvironment has different functions on different cells to control the banlance of tolerance or immunity in liver physiologically or pathophysiologically.