Article Text

Gut microbiota and Toll-like receptors set the stage for cytokine-mediated failure of antibacterial responses in the fibrotic liver
  1. Christian Kuntzen,
  2. Robert F Schwabe
  1. Department of Medicine, Columbia University, New York, New York, USA
  1. Correspondence to Dr Robert F Schwabe, Department of Medicine, Columbia University, Room 926, 1130 St. Nicholas Avenue, New York, NY 10032, USA; rfs2102{at}

Statistics from

The gut–liver axis is increasingly recognised as a key contributor to chronic liver disease. A failing gut barrier contributes to increased bacterial translocation, which results in an elevated risk of bacterial infection and a chronic inflammatory state that may promote the progression of chronic liver disease and the development of long-term complications such as fibrosis and HCC.1 ,2 The most important clinical consequence of increased translocation is acute bacterial infection, a common cause of hospital admissions and a major contributor to morbidity and mortality in patients with cirrhosis. Moreover, bacterial infections can lead to acute decompensation, often triggering acute-on-chronic liver injury.3 On top of a leaky gut, patients with liver cirrhosis have severe defects in the innate immune system, affecting macrophages, neutrophils and the complement system.4 The liver itself represents an important immunological organ and is the first target of gut-derived bacteria, bacterial pathogen-associated molecular patterns (PAMPs) and food products after they enter the circulation. For this reason, 80% of the body's resident macrophages are found in the liver, where they act as an important component of a firewall, that protects the body from infection from circulating bacteria.5 Following infection, a large proportion of circulating bacteria are phagocytosed by Kupffer cells rather than in the spleen; ablation of macrophages severely hampers clearance of circulating bacteria.5 Together, these findings suggest that the liver rather than the spleen is the main organ involved in the clearance of circulating bacteria. Importantly, the presence of liver fibrosis severely hampers the clearance of circulating bacteria.5 However, the mechanisms that impair the clearance of circulating bacteria in the fibrotic liver remain largely unknown. Better knowledge of the underlying pathways may lead towards novel therapeutic targets for patients with advanced liver disease.

In this issue, Hackstein et al6 found that the presence of liver fibrosis—induced either by bile duct ligation (BDL) or carbon tetrachloride (CCl4) injection—drastically hampered the ability of mice to respond to infection with the intracellular bacterium Listeria monocytogenes. Mice with liver fibrosis displayed up to 25-fold increased bacterial titres and high mortality, whereas the majority of non-fibrotic control animals cleared the infection and survived.6 While there were increased titres of CD11b+ cells in fibrotic livers, bacterial cultures showed decreased clearance of hepatic bacteria. This was explained by impaired phagocytic and bactericidal activity of hepatic monocytes and a failure to sufficiently increase granulocyte numbers after infection. Of note, bactericidal activity was not affected in the spleen in this setting, again emphasising the importance of the liver as relevant immunological organ in the clearance of bacterial infections. Up to this point, the findings by Hackstein et al. essentially confirm the key role of liver as an antibacterial firewall and a failure of this system in the fibrotic liver, with the difference that the authors used a Gram-positive intracellular pathogenic bacterium rather than Gram-negative extracellular commensals of previous studies.5 ,7 One important finding of Hackstein et al. is the discovery of a cytokine-mediated and potentially targetable pathway that mediates the failing antibacterial defence in the setting of liver fibrosis. The authors found a tonic increase of interferon β (IFNβ) in fibrotic livers and a subsequent IFNβ-mediated induction of interleukin 10 (IL-10) upon infection with intracellular bacteria, which were responsible for the suppressed antibacterial defence in the fibrotic liver.6 Accordingly, fibrotic mice deficient for interferon α-receptor (IFNAR) displayed reduced IL-10 levels after Listeria infection and were protected from Listeria-induced death, but not from BDL-induced fibrosis. Importantly, antibody-mediated neutralisation of IFNAR or IL-10 receptor promoted clearance of L. monocytogenes and prevented infection-associated mortality suggested that targeting these pathways in patients might also have beneficial effects on antibacterial defences (figure 1). Another key novelty of the current study is the discovery of a gut microbiota–Toll-like receptor (TLR)-mediated mechanism through which the leaky gut impairs antibacterial responses in the liver in the setting of fibrosis. In experiments with germ-free mice and with macrophages from TLR-deficient mice, the authors show an important role of gut-derived PAMPs in the tonic increase of IFNβ levels after bile duct ligation.6 Moreover, hepatic CD11b+ cells from bile duct-ligated germ-free mice showed an ameliorated phagocytic capacity in comparison with those from their specific pathogen-free counterparts. Of note, infection with the intracellular bacterium L. monocytogenes further increased the levels of IFNβ and IL-10, whereas infection with extracellular bacteria such as Escherichia coli, or a L. monocytogenes mutant that fails to invade the cytosol, did not. These data suggest that the presence of tonic IFNβ in combination with further exacerbation of IFNβ following infection with intracellular bacteria—most likely through cytosolic pattern recognition receptors (PRRs)—is key to the impaired antibacterial response. IL-10, a well-characterised anti-inflammatory cytokine, downregulated the levels of several other cytokines with known roles in antibacterial immunity, such as IFNγ, interleukin 12 (IL-12) and interleukin 1β (IL-1β)—all of which were restored when either IFNAR or IL-10 was blocked. Key findings of the study were confirmed in patients, showing a significant increase in IFNβ levels in patients with cirrhosis as well as an increased production of IL-10 by monocytes from patients with cirrhosis after infection with the intracellular bacteria, L. monocytogenes, Legionella pneumophila, Mycobacterium avium or Salmonella typhimurium.

Figure 1

Pathways mediating the failing antibacterial response in the fibrotic liver and therapeutic approaches targeting these pathways. In the setting of liver fibrosis, increased translocation of pathogen-associated molecular patterns (PAMPs) from a leaky gut results in activation of several Toll-like receptors (TLRs) in hepatic macrophages. TLR activation on macrophages leads to tonic upregulation of IFNβ and most likely minor upregulation of IL-10 and antibacterial cytokines IL-1β, IL-12 and IFNγ. Infection with intracellular bacteria such as Listeria monocytogenes amplifies IFNβ production in hepatic macrophages through activation of cytosolic pattern recognition receptors (PRRs). This leads to a drastic increase of IL-10 and suppression of antibacterial cytokines IL-1β, IL-12 and IFNγ, resulting in a failing antibacterial defence. Selective gut decontamination, for example, by non-absorbable antibiotics such as Rifaximin, pharmacological improvement of the gut barrier or blockade of type I IFN or IL-10 signalling could interrupt the signals that lead to failing antibacterial defences in the fibrotic liver and potentially reduce the occurrence of bacterial infections in patients with liver fibrosis. FXR, farnesoid X receptor.

It is important to discuss the findings and implications of Hackstein et al. in a clinical context. L. monocytogenes, a Gram-positive intracellular bacterium is a relatively rare cause of infection with an annual incidence of 0.29/100 000 in the USA. However, there is a higher incidence in patients with immunodeficiency from various causes.8 Accordingly, patients with cirrhosis have a >100-fold increased risk for listeriosis compared with the general population.8 Moreover, the incidence of infections with Mycobacterium tuberculosis, another intracellular bacterium, is increased in cirrhotics.9 However, infections with M. tuberculosis were relatively rare at 169/100 000/year. Based on the rare incidence of L. monocytogenes in the general population, a >100-fold higher incidence in cirrhotics still means that this infection is rare among cirrhotics. Besides teaching us about the role of the gut microbiota–TLR–IFNβ–IL-10 pathway in facilitating infections with intracellular bacteria, the study from Hackstein et al. may also help us to better understand the key role of the liver as an antibacterial firewall and the mechanisms that erode this role during fibrosis. As such, the most common bacterial infections in cirrhosis are caused by Enterobacteriaceae and non-enterococcal Streptococci, with characteristic complications such as spontaneous bacterial peritonitis, urinary tract infections, pneumonia and sepsis.3 Although the authors demonstrated that extracellular bacteria such as E. coli did not strongly exacerbate IFNβ and IL-10 levels in the setting of fibrosis, it is conceivable that even the observed moderate increase in IFNβ contributes to the failing antibacterial response in cirrhotic livers. If this were the case, targeting (i) the gut microbiota, (ii) the leaky gut, (iii) IFNβ or (iv) IL-10 might be beneficial in preventing bacterial infections—both with intracellular and with extracellular bacteria—in cirrhosis (figure 1). Data from a recent study, showing protection against Gram-negative sepsis by blockade of type I interferon signalling,10 are encouraging. A second key question is whether infection with intracellular bacteria or events that mimic this (e.g., activation of intracellular PRRs) could contribute to failing defences against extracellular bacteria. In such a scenario, even a clinically silent infection with intracellular bacteria (or pathophysiological events activating similar signals in monocytes) could trigger severe infections with extracellular bacteria. In summary, it will be important to determine (i) whether the failure of the hepatic antibacterial firewall against extracellular bacteria5 also involves the described IFNβ-IL-10-IFNγ/IL-12/IL-1β signalling cascade and (ii) to test whether clinically silent infections with intracellular bacteria may be a trigger for severe infections with extracellular bacteria in patients.

View Abstract


  • Funding National Institute on Alcohol Abuse and Alcoholism (grant no. U01AA021912).

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles