Apoptosis or programmed cell death is a common feature of chronic liver diseases, the result of ongoing toxic stimuli, culminating in the death of hepatocytes. During apoptosis, cascades of proetolytic and nucleolytic enzymes disassemble the cells, leading to DNA cleavage. The end-products of this process are the double-membrane surrounded apoptotic bodies, containing cleaved DNA and proteolytic fragments. The apoptotic bodies are subsequently cleared by phagocytosis.1, 2 Phagocytosis is an effective and rapid way of eliminating dead cells, and through the production of anti-inflammatory cytokines, maintaining homeostasis.2 In contrast to apoptosis, necrotic cell death is a more random process resulting from chemical injury or hypoxia. During apoptosis or necrosis, DNA from dying cells could be released to the extracellular space. This occurs if the phagocytic system is overwhelmed during acute tissue injury, or when the function of the phagocytic system is impaired (e.g., in the case of autoimmune diseases), and secondary necrosis develops.3 Extracellular DNA then could form immune complexes and also be found as circulating DNA in the serum.4, 5

The unmethylated extracellular cytidine-phosphate-guanosine (CpG)-DNA motifs are of particular significance. Bacterial or viral CpG DNA has long been recognized to have an immune stimulatory function, and now it is becoming clear that mammalian CpG DNA, originating from apoptotic cells, has a similar role. Recently it has been reported that CpG-rich extracellular DNA is produced in states of autoimmunity systemic lupus erythematosus [SLE] and the production is linked to excessive apoptosis and impaired phagocytic clearance.6, 7 CpG DNA is recognized by the TLR9, a member of the Toll-like receptor (TLR) family. TLR9 is an intracellular receptor therefore CpG DNA has to be internalized for receptor binding to occur. TLR9 is essential in recognizing microbial CpG rich DNA sequences,8 and now there is evidence that it recognizes mammalian DNA, as well. TLR9 signaling is activated by immune complexes in patients with lupus nephropathy,9, 10 and plays a role in the immune complex-mediated activation of B cells and plasmocytoid dendritic cells.11

What is the relevance of CpG DNA in the pathogenesis of liver diseases? Data providing direct evidence for the presence and role of CpG DNA during acute or chronic liver injury are emerging. It has been described that in patients with primary sclerosing cholangitis who have biliary epithelial cell antibodies, TLR9 is up-regulated, and it is postulated that biliary epithelial cell antibodies cause an exaggerated activation of the innate and adaptive immune system in response to the activation of TLR9.12 Furthermore, CpG-DNA was shown to induce secretion of antimitochondrial antibodies in PBC patients, and also up-regulate TLR9 expression in B cells.13 This suggests that hyperresponsiveness of B cells in primary biliary cirrhosis accelerates B cell–mediated autoimmunity. An indication that CpG DNA plays a role in acute liver injury is provided by Yi et al.14 In this study, treatment of D-galactosamine-sensitized mice with CpG-DNA caused massive liver injury by inducing caspase 9-mediated apoptosis of hepatocytes. The effects of CpG DNA were mediated by TLR9/MyD88 activation. As primary heptocytes did not express significant amount of TLR9, CpG-DNA-mediated cytokine production by macrophages/dendritic cells was thought to be the cause of liver injury.

As to how CpG DNA/TLR9-mediated signaling plays a role in chronic liver diseases, and in particular, what is the contribution of this pathway to liver fibrogenesis, is still unknown. During chronic liver injury apoptotic cells are cleared by both professional and nonprofessional phagocytes. It was recently shown that hepatic stellate cells (HSC), which are at the center of the fibrogenic process, engulf apoptotic bodies, which triggers a profibrogenic response with upregulation of transforming growth factor (TGF)-β1 and procollagen α 1 (I) expression.15, 16 It is also possible that HSC engulf necrotic debris, and uptake DNA from apoptotic cells as this is suggested by Watanabe et al.17 Since HSC acquire a migratory phenotype upon transdifferentiation, it is of great interest, how they reach and recognize the area of cell injury. Platelet-derived growth factor (PDGF) is a major motogenic cytokine for stellate cells, and several studies described its crucial role in protrusion, lamellipodia formation and actin-myosin cytoskeleton rearrangement.18, 19 Recently, exciting new studies have been emerging elucidating the stop signals for migrating HSC. For instance, adenosine release from areas of apoptosis was shown to strongly inhibit PDGF-induced HSC chemotaxis.20 In the current issue of HEPATOLOGY Watanabe et al. made an important observation that CpG DNA from apoptotic hepatocytes inhibits stellate cell chemotaxis via TLR9 signaling.17 Exposure of HSC to apoptotic DNA from hepatocytes or CpG DNA was also found to up-regulate TGF-β1 and collagen-1 expression involving a TLR9 mediated pathway, as TLR9 antagonists blocked the up-regulation. This is the first time that TLR9 expression is described both in a HSC line and in primary stellate cells. According to current concepts TLR9 is expressed mainly by plasmacytoid dendritic cells therefore its presence in HSC points to the fact that HSC may have immune modulatory and antigen-presenting roles. Indeed, according to recent studies HSC were described as intrahepatic antigen-presenting cells, eliciting T cell responses to protein and lipid antigens.21

How does CpG-DNA/TLR9 mediated signaling result in upregulation of profibrogenic genes? Generally, upon recognizing CpGs, intracellular TLRs are known to recruit an adaptor protein, myeloid differentiation marker 88 (MyD88), IL-1 receptor–associated kinase (IRAK), and tumor necrosis factor receptor–associated factor 6 (TRAF6). These adaptors in turn mediate the activation of the jun NH2-terminal kinase (JNK), nuclear factor (NF)-κB, p38, extracellular signal–related kinase 1/2 (ERK 1/2), and phosphoinositide 3–kinase signaling pathways leading to activation of inflammatory target genes.22 Whether TGF-β and collagen up-regulation in HSC results directly from these pathways, or requires other intermediaries, is not known. It is described however, that TLR9 induction by CpG DNA contributes to the progression of interstitial fibrosis with increased collagen-I mRNA expression in MRL (Murphy Roths Large)lpr/lpr mice, a model for lupus nephritis.23 In this model, TLR9-mediated CCL5/RANTES signaling was thought to play a role in the progression of renal fibrosis.

Also of interest is the mechanism by which CpG DNA is internalized by HSC and its subsequent recognition by the TLR9 receptor. Most of the knowledge acquired in this field is based on studies conducted in macrophages and dendritic cells. It is interesting to postulate that HSC may exhibit many macrophage-like characteristics, such as a capacity to phagocytose, antigen-presentation, and now recognition and uptake of CpG DNA and subsequent TLR9 signaling. Based on current models, TLR9 is initially localized to the endoplasmic reticulum (ER).24 CpG DNA is internalized via a clathrin-dependent endocytic pathway then rapidly transported to a lysosomal compartment, where TLR9 is recruited from the ER. CpG DNA then directly binds TLR9 in the presence of MyD88, and signaling events follow (Fig. 1). The high mobility group box (HMGB) proteins have been recently described to functionally interact with class A CpG DNA and the resulting complex activates plasmacytoid dendritic cells via TLR9.11 HMGB1 DNA is released from cells undergoing necrosis, but not apoptosis, or from cells exposed to inflammatory cytokines. Whether HMGB1 DNA is released from dying hepatocytes and also participates in the activation of HSC, remains to be determined.

thumbnail image

Figure 1. Schematic of TLR9 signaling in hepatic stellate cells. See text for details. AB, apoptotic bodies; CpG, cytidine-phosphate-guanosine; TLR9, Toll like receptor 9; PDGF, platelet derived growth factor; MyD88, myeloid differentiation marker 88; TGF-β, transforming growth factor beta.

Download figure to PowerPoint

The concept that CpG DNA in HSC may be recognized by specific receptors is novel, and would merit further analysis regarding the mechanism of its recognition, uptake and binding to TLR9. An interesting alternative hypothesis is that CpG DNA is engulfed by HSC in the form of apoptotic bodies. As apoptotic DNA is rich in CpG DNA and processed by phagolysomes, then lysosomes; it can induce TLR9 recruitment and binding of CpG DNA, thus initiating signaling pathways. For instance, TLRs were shown to mediate the induction of genes involved in the recognition of apoptotic cells such as the ATP-binding-cassette transporter 1 (ABC1).25

The authors provide evidence for an inhibitory effect of CpG DNA/TLR9-mediated signaling on PDGF-induced chemotaxis of HSC. Activation of TLR9 resulted in an inhibition of IP3-mediated signaling and a decrease in the cytosolic Ca++ concentration, which correlate with reduced motility of HSC. Studies assessing matrix metalloproteinase (MMP) activation in response to a TLR9 agonist may also be relevant in the future as TLR9 was shown to activate MMP13 activity, and this also could have a direct consequence on HSC migration.

TLR9 activation in HSC acts as a double-edged sword: it provides a stop signal for migrating activated HSC as soon as they “sense” apoptotic DNA, and further induces their activation and differentiation evidenced by the up-regulation of profibrogenic genes. These findings are well-supported by the in vivo experiments where TLR9−/− mice had a decrease in collagen deposition and HSC activation based on a CCl4 model of fibrosis. As TLR9 agonists may have untoward effects on the liver,14 studies in the future may focus on application of specific oligodeoxynucleotides sequences (ODNs) that could inhibit TLR9 activation in the liver.


  1. Top of page
  2. References