Dendritic cells (DCs) are particular hematopoietic cells that link the innate and adaptive part of immune responses. DCs, regularly found in lymphoid and nonlymphoid tissues, are specialized cells for presenting antigens in conjunction with major histocompatibility complex (MHC) class II molecules to T cells for initiating antigen-specific immune responses. Although DCs share certain characteristics, such as intracellular processing of phagocytized peptides and proteins for antigen presentation, migratory properties (toward the draining lymph node), and cytokine production, several functionally distinct DC subsets have been unraveled in mice and humans. However, most of these DC functions have been uncovered in infectious or autoimmune disease models in typical lymphoid organs, such as spleen or lymph nodes, and relatively little is known at present on the possible roles of DC populations in the liver. Moreover, given that many conditions of liver inflammation, such as nonalcoholic steatohepatitis (NASH), are classically considered noninfectious inflammatory reactions and are certainly not directed against a single antigen, unlike immune responses against viral proteins in viral hepatitis, the relevance of DCs for regulating sterile liver inflammation and fibrosis is even less clear. Nevertheless, independent studies had reported on the accumulation of myeloid cells with DC characteristics in experimental models of toxic and cholestatic liver diseases.[3-6] In this issue of Hepatology, Henning et al. explored the potential role of DCs in regulating hepatic inflammation and fibrogenesis in NASH (Fig. 1).
Upon induction of experimental steatohepatitis by feeding a methionine-choline deficient (MCD) diet over 6 weeks, the number of DCs, as defined by positive staining for the leukocyte marker, CD45, the mouse DC marker, CD11c, and MHC class II molecules, markedly increased in injured livers. In comparison to normal livers, NASH-associated DCs showed a more mature phenotype with respect to expression of costimulatory molecules, produced increased levels of cytokines, and displayed an enhanced capacity to activate antigen-specific CD4 T cells, but not CD8 T cells, when isolated from steatotic murine livers. From these data, one could have speculated that these DCs promote inflammatory reactions in NASH. To test the functional role of these cells in vivo, the researchers used a mouse model to deplete these cells continuously during NASH progression by administration of diphtheria toxin (DT) to bone marrow chimeric mice, in which all hematopoietic cells carried the diphtheria toxin receptor (DTR) on CD11c-expressing cells. Surprisingly, depletion of CD11c-expressing cells augmented intrahepatic inflammation, especially the activation of neutrophils, Kupffer cells, and inflammatory monocytes in injured livers, increased the number of apoptotic cells, and accelerated hepatic fibrosis. The researchers provide some indirect data supporting that DCs may limit inflammation in NASH liver by clearing necrotic cellular debris and apoptotic bodies, which would fit well with the observed increased number and activation of innate immunity in DC-depleted mice (Fig. 1). Furthermore, when DCs were abrogated during recovery from MCD diet, DC depletion delayed the resolution of experimental NASH, which is well in agreement with an earlier study demonstrating that DCs facilitate fibrosis regression in a model of sterile toxic liver injury by carbon tetrachloride.
The elegant study by Henning et al. strongly indicates beneficial functions of hepatic DCs in limiting fibroinflammatory reactions in the steatotic environment (Fig. 1), which might thus represent an attractive target for future therapeutic strategies. Nevertheless, before developing DC-based immunomodulatory therapies for NASH, it is important to keep some shortcomings of the model in mind when interpreting the data. The researchers depleted CD11c-expressing cells; this molecule is an accepted DC marker in mice, but not for human DCs, which may hamper translating these results into human disease, especially because DC subsets were not targeted in a specific manner by this approach. Moreover, CD11c expression is not exclusively confined to DCs in mice, because several important immune cells, including natural killer cells, and macrophage subsets regularly express CD11c in injured mouse liver. Third, by using this depletion strategy, all CD11c-expressing cells in the body were effectively depleted, which leaves the possibility that some of the favorable net effects of DCs on liver inflammation may not be conducted by hepatic DCs, but by DCs in other compartments, as previously shown for tumor or sepsis models.[10, 11] Fourth, it has been recently described that neutrophilia can be a “side effect” of DC ablation upon DT injection in this specific model of CD11c-DTR transgenic mice. Although the researchers attempted to control for this concern by using the bone marrow chimeric animals, it raises the possibility of confounding effects on inflammatory responses related to the model, but not to DC depletion.
However, the overall anti-inflammatory and antifibrotic function of hepatic DCs in the NASH model was accompanied by a rather activated DC phenotype with high cytokine secretion and efficient T-cell stimulatory capacity ex vivo. Similarly, it had been reported that DCs isolated from experimental nonalcoholic fatty liver disease (NAFLD; high-fat and high-calorie model) showed impaired function in antigen processing, despite that these cells produced higher levels of inflammatory cytokines and showed increased T-cell proliferation. One possible explanation is that the lipid content within DCs severely reduces their capacity to process antigen without affecting the expressions of MHC-II and costimulatory molecules, because it has been observed for obesity-related cancer. Further experiments should address the functionality and antigen processing of hepatic DCs dependent on the intracellular lipid levels.
Nevertheless, DC accumulation in experimental NASH appeared to down-modulate several innate immune cell components, including neutrophils and macrophages. This observation is consistent with previous data supporting that DC migration to injured liver (e.g., in Propionibacterium acnes–induced granulomatous liver diseases) is critical to limit inflammation and restore the hepatic microenvironment to its normal state. Hepatic DCs thus appear to share some functional similarities to myeloid-derived suppressor cells (MDSCs), which have been identified to suppress immune responses in conditions of malignant diseases or organ transplantation. Therefore, it is possible that hepatic DCs with such MDSC-like phenotype may down-regulate fibrogenesis, but favor the development of hepatocellular carcinoma (HCC) (Fig. 1). MDSCs have been linked to HCC progression, and NASH is recognized as an increasingly important predisposition for HCC, both in cirrhotic and noncirrhotic liver. Thus, the clear beneficial role of hepatic DCs in NASH-associated fibrosis by down-modulating innate immune cell components should be further explored with respect to their effect on HCC, because the MDSC-like property of (lipid-laden) DCs could favor tumor development in NASH.
Frank Tacke, M.D., Ph.D.1
Hiroyuki Yoneyama, M.D., Ph.D.2
1Department of Medicine III University Hospital Aachen Aachen, Germany
2Stelic Institute & Co. Tokyo, Japan