Introduction

Nonalcoholic fatty liver disease (NAFLD) is one of the most common chronic diseases of the liver, even in pediatrics. It nowadays reaches a prevalence of up to 40 % in the general population and up to 13 % in children [1]. These prevalence data reflect the parallel increase of the risk factors of this disease, such as obesity and type 2 diabetes mellitus.

NAFLD ranges from simple fat accumulation (steatosis) to inflammation and fibrosis [nonalcoholic steatohepatitis (NASH)]. The progression of NAFLD to the more severe form, NASH, is linked to various genetic and environmental factors [2•]. The progression of disease is a crucial aspect, since the evolution of fatty liver to fibrotic stages may be the prelude to cirrhosis and hepatocellular carcinoma (HCC) [3], even if in children only one case of steatosis-dependent HCC has been reported in the absence of a fibrotic background [4]. Plenty of research has focused on the interaction between the liver and gut, since the so-called “gut–liver axis” appears to play a major role among the factors leading to disease progression [5, 6•].

Normally the intestinal epithelium allows the absorption of nutrients while also functioning as a barrier, which prevents antigens and pathogens from entering the mucosal tissues and potentially causing disease. The intestinal tract is inhabited by 1014 microbes [7], and it has become evident that they are involved in molecular cross-talk with the intestinal epithelium and affect the intestinal barrier function [8]. An important component of the intestinal barrier is the intercellular junctional complex, crucial for the maintenance of barrier integrity. Tight junctions (TJs) constitute a multifunctional complex that forms a seal between adjacent epithelial cells near the apical surface [9]. TJs seal the paracellular space between epithelial cells, thus preventing paracellular diffusion of microorganisms and other antigens across the epithelium. They are not a static barrier, but highly dynamic structures that are constantly being remodeled because of interactions with external stimuli, such as food residues and pathogenic and commensal bacteria. In this way, TJs can regulate the entry of nutrients, ions and water, but restrict pathogen entry and maintain the barrier function of the epithelium.

NAFLD is associated with increased intestinal permeability (IP), since TJs appear to go through a disruption process making the intestine in these liver patients “leaky.” In in vitro and animal models of NAFLD, increased IP and alterations of gut microbiota have been shown to augment the exposure of the liver to gut-derived bacterial products. Increased serum levels of these products [for example lipopolysaccharides (LPS)] may cause low-grade endotoxemia, especially in the portal system and in the liver. Endotoxemia stimulates innate immune receptors, which activate signaling pathways involved in liver inflammation and fibrogenesis [10•].

In this setting, it is not surprising that probiotics reduce hepatic injury and inflammation in ob/ob mice [11], suggesting that modulation of the gut microbiota by probiotics may represent a feasible approach for the prevention or treatment of NASH.

The purpose of this review is to provide an overview of the current knowledge on the pathophysiology of the cross-talk between the gut and the liver in the pathogenesis of NAFLD and its progression to NASH, trying to open new treatment horizons in this field.

Role of Intestinal Permeability in the Gut–Liver Cross-Talk

It is now well known that in case of small bowel bacterial overgrowth (SBBO) and increased IP, the hepatobiliary system is inevitably exposed to a high level of these bacterial products and is thus able to activate the receptors on cell surfaces and to initiate the cascade of proinflammatory signal transduction leading to liver inflammation and fibrosis. In patients with NAFLD and correspondents in animal models, both alterations are demonstrated.

In a case-control study, Wigg et al. [12] showed that patients with NASH had a higher prevalence of SBBO, as well as blood levels of tumor necrosis factor alpha (TNF-α), compared to controls. Another case-control study by Miele et al. [13] has given particularly significant results. This study was able to confirm not only that NAFLD is associated with SBBO, but also that it is associated with an increased IP because of a process of destruction of the TJs of the intestinal wall, which in turn is associated with the severity of hepatic steatosis. Miele et al. [13] studied 35 consecutive patients with biopsy-proven NAFLD, 27 with untreated celiac disease (as a model of IP), and 24 healthy volunteers. They assessed the presence of SBBO by glucose breath testing, IP by means of the urinary excretion of 51Cr-ethylene diamine tetraacetate (51Cr-EDTA) test, and the integrity of TJs within the gut by immunohistochemical analysis of zona occludens-1 (ZO-1) expression in duodenal biopsy specimens. Patients with NAFLD had significantly increased gut permeability (compared with healthy subjects; p < 0.001) and a higher prevalence of SBBO, although both were lower than in the untreated celiac patients. In patients with NAFLD, both IP and the prevalence of SBBO were correlated with the severity of steatosis, but not with the presence of NASH. Our group has recently conducted a case-control study on a pediatric population [14], investigating the prevalence of altered IP in children with biopsy-proven NAFLD and studying its potential association with the stage of liver disease. LPS blood levels were also evaluated. IP in children was measured using the lactulose-mannitol bowel permeability test [15]. In line with previous findings, IP was significantly higher in children with NAFLD compared to controls (p < 0.05). Within the NAFLD group, IP was increased in children with NASH compared to those with steatosis only (p < 0.05) and was correlated with the presence and stage of portal inflammation (p = 0.02), fibrosis (p = 0.0002) and ballooning of hepatocytes (p = 0.003). Moreover, circulating LPS levels were higher in children with steatohepatitis (p < 0.05). IP and LPS blood levels increased in patient with NAFLD and were correlated with the severity of disease. Therefore, qualitative or quantitative changes in the intestinal bacterial flora and in their toxic products may lead to disruption of the intestinal barrier, bacterial translocation across the gut mucosa and the development of portal endotoxemia.

Why a Link Between the Liver and the Gut?

The liver is a very peculiar organ, since it receives blood from both the portal vein and the hepatic artery. The blood via the portal vein, draining the mesenteric veins, contains not only products derived from food digestion, but also from the bacteria that physiologically colonize the gut. In fact, it has been suggested that the portal vein serves as a “super highway” from the intestine to the liver [16••].

The portal system carries about 70–75 % of the total hepatic blood flow and releases bacterial products such as the LPS, bacterial DNA and peptidoglycan into the liver. All these molecules belong to distinct classes of endogenous signals, called pathogen-associated molecular patterns (PAMPs) or damage-associated molecular patterns (DAMPs), that trigger the host immune response by the activation of Toll-like receptors (TLRs) [17]. The protective role of TLRs against numerous microbes is mostly achieved by activating some of the dominant signaling cascades such as the nuclear factor kappa B (NFκB) transcriptional pathways, which are referred to as proinflammatory since they promote immune cell recruitment [17].

It has been demonstrated that in NASH patients, even low blood levels of microbial products—endotoxemia—might activate TLR pathways in the liver. In fact, TLRs are a class of receptors that recognize structurally conserved microbe-derived molecules and are expressed on Kupffer cells, biliary epithelial cells, hepatocytes, hepatic stellate cells (HSC), epithelial cells and dendritic cells of the liver [6•]. They are part of the mucosal immune system and have the major role in detecting and clearing most food-borne pathogens as well as preventing harmful bacteria from harming beneficial bacteria and host tissues.

While immune cell recruitment exerts a containment action against pathogens, it can also result in host tissue damage. It is well known that cytokines that mediate the inflammatory responses such as TNF-α also have a variety of effects on differentiation pathways; thus, there is a potential for microbial products to have broad effects on host phenotype. The latter seem to happen in the development and progression of fatty liver disease. In fact, some studies have shown that patients with NAFLD have an increased blood level of LPS, especially if affected by NASH or initial stages of liver fibrosis [6•, 18]. Moreover, levels of bacterial endotoxins seem correlated to the levels of proinflammatory cytokines, such TNFα. Yang et al. [19] were the first researchers to observe that LPS is able to promote the evolution of NAFLD in NASH in animal NAFLD models through the activation of TNFα. Wigg et al. [12] also confirmed these findings in humans in a case-control study: patients with NASH had higher a prevalence of SBBO as well as higher blood levels of TNFα compared to controls. This demonstrates the cross-talk between the gut and liver in NAFLD.

The Cross-Talk Between the Liver and Gut in NAFLD and Its Related Fibrosis

TLRs are present on the cells of innate immunity and are able to recognize highly conserved motifs in PAMPs and DAMPs. TLRs, acting as immune sensors of PAMPs and DAMPs, initiate an adaptive immune response and a signaling cascade leading to activation of proinflammatory genes. The most studied PAMP is LPS, a component of the gram-negative bacterial cell membrane, the active component of endotoxin. The latter binds to LPS-binding protein, which subsequently binds to CD14.

The LPS–LPS-binding protein (LBP)–CD14 complex activates TLR-4, present on Kupffer cells, triggering an essential inflammatory cascade. Because of the unique “blood link” between the liver and the gut described above, the liver is constantly exposed to TLR ligands. Normal livers do not show signs of inflammation because of low expression of TLR-4 and its adaptor molecules in the liver and the ability to modulate TLR-4 signaling by liver tolerance. But under pathologic conditions such as the increased IP that is seen in NAFLD, altered TLR-4/LPS signaling plays an important role in the pathogenesis and progression of chronic fatty liver disease.

Of all the 13 TLRs known, TLR-2, TLR-4, its coreceptor CD14 and TLR-9 have been well studied in NAFLD. TLR4 specifically is the receptor of LPS. Szabo and coworkers investigated the role of TLR-2 and TLR-4 polymorphism on liver damage and on cytokine induction in a methionine–choline-deficient (MCD) diet-induced model of NASH [20]. They found that in TLR-2−/− mice there is an increase in liver injury associated with NASH, which may suggest a protective role for TLR-2−/− mediated signals in liver injury [20]. Similarly, Rivera et al. [21] found that hepatocellular damage was notably more severe and the TNF-α level was more elevated in TLR-2−/− mice. Possibly the TLR-2 deficiency exacerbates NASH by altering the signaling via the TLR-4 pathway, whereas its presence may play a protective role against the induction of NASH.

Chronic liver injury leads to the development of hepatic fibrosis because of an inflammatory trigger and the consequent increased accumulation of extracellular matrix in the liver [22]. It is now clear that if one hand is made by the Kupffer cells, which are those initiating fibrogenesis by secreting proinflammatory and profibrogenic cytokines, the HSCs are the predominant source of extracellular matrix production in the fibrotic liver [23].

Noteworthily, abundant data demonstrate that LPS is elevated in experimental models of hepatic fibrosis [24–26] and in patients with cirrhosis [27–29]. It is believed that alterations of the intestinal microbiota and a failure of the intestinal mucosal barrier cause increases in bacterial translocation and LPS levels, especially in later stages of hepatic fibrosis and cirrhosis [30–35]. Recent studies on mice models have demonstrated the crucial role for the LPS–TLR4 pathway in hepatic fibrogenesis. In fact, TLR4 is expressed on two key mediators of hepatic fibrogenesis: Kupffer cells and HSC [2•, 36, 37]; although Kupffer cells express the highest levels of TLR4 in the liver and are considered a prime target of LPS, TLR4 expressed on quiescent and activated HSC is the main mediator of fibrosis. It has been demonstrated that TLR4-chimeric mice display an important reduction in hepatic fibrogenesis [24]. LPS directly targets HSC in vivo to upregulate chemokines and attract Kupffer cells. At the same time, TLR4 activation induces a downregulation of the TGFβ pseudoreceptor Bambi on HSC. These two mechanisms work hand in hand to promote the activation of HSC by Kupffer cell-released TGFβ and subsequently hepatic fibrosis. In fact, HSCs [38] have a TLR4-mediated NF-κB activation in response to a fairly low concentrations of LPS. NF-κB is reported to be the predominant target through which TLR4 ligands promote fibrosis in the liver [39].

The crucial role of LPS is also supported by the finding that LBP-deficient as well as gut-sterilized mice also have a marked reduction of hepatic fibrosis [24, 40]. In addition to LPS, endogenous TLR ligands such as the two most known DAMPs, hyaluronan and high-mobility group box 1 (HMGB1), are also elevated in murine fibrogenesis [24]. Interestingly, a recent experimental study demonstrated that HMGB1 is able to induce proliferation and activation of HSC and to trigger their proinflammatory and profibrogenic phenotype by increasing the expression of monocyte chemoattractant protein-1 (MCP-1), enhancing the effect of TGFβ [41•].

Also the TLR9 recognizes fragments of bacterial DNA in intestinal derivation and is able to induce the production of proinflammatory cytokines; in fact TLR9 appears to be involved in the evolution of nonalcoholic fibrotic liver disease through the production of IL1β by the Kupffer cells. Indeed knockout mice for TLR9 have a reduced degree of steatohepatitis and fibrosis [6•].

Modulating the intestinal microflora and consequent endotoxemia levels through the use of probiotics and selective intestinal decontamination may open new ways to avoid NAFLD complications with hepatic fibrosis, cirrhosis and HCC.

Effects of Intestinal Microbiota Modulation on NASH

Probiotics are live commensal microorganisms that can modulate the intestinal microbiota. A first systematic review in 2007 [42] underlined the absence of randomized controlled trials on the administration of probiotics in patients with NAFLD. However, to date, several studies have been performed in both animal models and in humans, some of which are randomized controlled trials (see Table 1), showing encouraging results for the use of these formulations.

Table 1 Available RCTs with probiotics in nonalcoholic fatty liver diseases
Full size table

Changes in the intestinal microbiota through the use of probiotics seem to be able to vary NAFLD progression, limiting the progression of fibrogenesis in several animal models of NAFLD [43, 44]. Performing interventions on intestinal microbiota using probiotics has been demonstrated to modulate the expression of nuclear receptors, correcting insulin resistance in the liver and the adipose tissues and protecting against the development of NASH and progression to fibrosis.

VSL#3 is the most largely studied probiotic in the field. In particular, Velayudham et al. demonstrated VSL#3, a mixture of eight probiotic strains (Streptococcus salivarius subsp. Thermophilus, Bifidobacterium [B. breve, B. infantis, B. longum], Lactobacillus acidophilus, L. plantarum, L. casei, and L. delbrueckii subsp. Bulgaricus), exerted a protective action against fibrogenesis, but not on hepatic inflammation, in a MCD diet-induced mouse model of NAFLD [43]. In mousee models of genetic dyslipidemia (Apo-E-deficient mice) failing to develop NASH-like lesions on a standard diet, it has been shown that destrane sulfate sodium-induced intestinal inflammation and the consequently increased IP triggered the transition of steatosis to NASH and that these disorders were efficiently prevented by a therapeutic intervention with VSL#3 [45]. Similar results were also obtained in animal models of high-fat-diet-induced liver disease that was attenuated by VSL#3 treatment [46].

More recently, the MIYAIRI 588 strain of C. butyricum has been reported to induce a decrease in hepatic fibrous deposition in rat models of choline-deficient/l-amino acid-defined (CDAA)-diet-induced NAFLD [43].

In intestinal tissues in a rodent model of colitis, it was demonstrated that probiotics can modulate IP and correct the inflammation-driven metabolic dysfunction [46]. Therefore, it is conceivable that in theory probiotic administration might reset the “leaky gut” of NAFLD patients, offering an interesting approach to counteracting liver damage in NAFLD even though no well-designed clinical trials with effective results are available yet [47].

Conclusions

The gut and liver are anatomically strictly linked to one another. Alterations in the gut barrier through a TJ disruption process whose mechanisms are not entirely clear break the established balance between the two organs. Consequent to destruction of the IP integrity, harmful molecules are released in the portal vein blood and act against the liver to activate inflammation and fibrogenesis. TLRs appear to play a crucial role at this point, since their signaling cascade induces the production and release of proinflammatory signals in Kupffer cells and the expression of a profibrogenic pattern in HSCs. The host inflammatory response of Kupffer cells and HSC activation via TLRs therefore develop into a vicious cycle that, in a dangerous liaison between the gut and liver, causes low-grade hepatic inflammation and damage, inducing the shift from simple steatosis to a more aggressive NASH coupled with fibrosis (Fig. 1).

Fig. 1
figure 1

The role of the gut in the pathogenesis of NASH

Full size image

As the diet-dependent alteration of the composition of the intestinal microbiota is at the root of this process, therapeutic approaches, including probiotic-based treatments, seem able to restore the physiological microbioma. Interventions on the microflora using probiotics appear to be promising against NAFLD-related tissue damage. Despite early experimental studies, no clinical evidence to support the use of any such probiotics is available. Hence, further studies are needed to understand whether probiotics may realistically open a new page in the treatment of NAFLD-related damage or at least may represent a concrete support to increase the effectiveness of already proven targeted drugs.