Pro‐inflammatory signalling and gut‐liver axis in non‐alcoholic and alcoholic steatohepatitis: Differences and similarities along the path

Abstract Non‐alcoholic fatty liver disease (NAFLD) and alcohol‐associated liver disease (ALD) represent a spectrum of injury, ranging from simple steatosis to steatohepatitis and cirrhosis. In humans, in fact, fatty changes in the liver, possibly leading to end‐stage disease, were observed after chronic alcohol intake or in conditions of metabolic impairment. In this article, we examined the features and the pro‐inflammatory pathways leading to non‐alcoholic and alcoholic steatohepatitis. The involvement of several events (hits) and multiple inter‐related pathways in the pathogenesis of these diseases suggest that a single therapeutic agent is unlikely to be an effective treatment strategy. Hence, a combination treatment towards multiple pro‐inflammatory targets would eventually be required. Gut‐liver crosstalk is involved not only in the impairment of lipid and glucose homoeostasis leading to steatogenesis, but also in the initiation of inflammation and fibrogenesis in both NAFLD and ALD. Modulation of the gut‐liver axis has been suggested as a possible therapeutic approach since gut‐derived components are likely to be involved in both the onset and the progression of liver damage. This review summarizes the translational mechanisms underlying pro‐inflammatory signalling and gut‐liver axis in non‐alcoholic and alcoholic steatohepatitis. With a multitude of people being affected by liver diseases, identification of possible treatments and the elucidation of pathogenic mechanisms are elements of paramount importance.


| INTRODUC TI ON
In the United States, nearly 35% of the population is obese [body mass index (BMI) >30] and another 6.3% is considered to have morbid obesity (BMI > 40 or more); the majority of these people have or may develop insulin resistance and/or metabolic syndrome, both of which are related to non-alcoholic fatty liver disease (NAFLD). 1,2 NAFLD affects 1%-20% of the US population, whereas non-alcoholic steatohepatitis (NASH) affects 1.1%-14% of US population. 3,4 NAFLD itself is not necessarily harmful; however, it can lead to NASH that is characterized by prolonged chronic liver inflammation and injury. NASH is one of the leading causes of liver cirrhosis in the United States together with alcoholic liver disease (ALD). Regarding the latter, nearly 50% of all cirrhosis related deaths are associated with chronic alcohol consumption. 5 Other types of ALD include fatty liver, alcoholic steatohepatitis (ASH) and cirrhosis. 6 Approximately, 20%-30% of subjects with chronic alcohol abuse may develop fatty liver, whereas 15% of patients may develop cirrhosis with 1% prevalence of hepatocellular carcinoma (HCC). 7  There is a translational link between NAFLD and ALD. NAFLD is a liver disease which includes fatty liver (NAFL) which is considered benign disease, steatohepatitis (NASH), which causes inflammation and accumulation of fat and scar tissue in the liver and can progress to advanced fibrosis, cirrhosis and its complication like HCC. 8 Although a similar condition can occur in people who abuse alcohol, NASH usually occurs in those who drink little to no alcohol.
ALD spectrum comprises simple steatosis, alcoholic hepatitis, and cirrhosis and HCC. There is significant overlap of the translational link between NAFLD and ALD, and clinical presentation depends upon the stage of liver disease. NAFLD patients are mostly asymptomatic and diagnosed to have fatty liver while undergoing routine health examination. 8 ALD requires significant history of alcohol intake which is supportive by serum and biochemical tests. In both NAFLD and ALD patients, liver enzymes are usually not significantly changed. Liver biopsy is required for diagnosis of NASH as it is a histological diagnosis and sometimes in alcoholic hepatitis for confirmation if diagnosis is uncertain. Some serum/ultrasound markers and prognostic scores have been developed for shunning liver biopsy in evaluation and treatment response of NASH and alcoholic hepatitis patients. 9 In a study involving 52 patients (26 with NASH and 26 with ALD), few parameters discriminated NASH from ALD. The most obvious factors were the nutritional status and features of visceral fat deposition. Histological rates of fibrosis and necrosis in NASH and ALD were similar, whereas patients with NASH exhibited higher levels of steatosis. The degrees of lipid peroxidation were similar, 10 despite the increased cholesterol serum levels in patients with NASH phenotypes. Although many obese people develop NASH, this disease has been also observed in non-obese, lean subjects; on the other hand, only 35% of all heavy drinkers develop cirrhosis. These observations suggest that other factors may contribute to human NASH and ASH onset/progression (Table 1). In this context, possible contributing factors include the following: (a) lipotoxicity; (b) mitochondrial dysfunction and oxidative damage; (c) innate immune system; and (d) changes in the gut-liver axis. Comparative pathogenesis of NASH and ASH as a function of different pathways involved and the role of the liver-gut axis in these events will be outlined in this review.

| FAT T Y LIVER
In normal conditions, even if the liver is involved in the management/disposal of an important traffic of lipids (mainly composed | 5957 GLASER Et AL. by triglycerides, and fatty acids), fat deposition in tissue does not occur. Low concentration of lipids in the graft is obtained by a fine balance between fat uptake/neo-synthesis and fatty acid oxidation plus lipoprotein assembly/disposal. 11 An imbalance in fat recruitment and its use in the liver determines an abnormal accumulation of lipids in hepatocytes thus resulting in fatty liver. 12 The characteristic morphological feature (by histology or with imaging techniques) of this condition is represented by liver steatosis. Hepatic steatosis is found in patients with NAFLD, NASH and ALD and is diagnosed when more than 5% of liver cells contain fat at microscopic examination. 13 However, there are several causes, with lower prevalence, that may determine fatty liver, including surgical, toxic, viral and inborn errors. In the course of steatosis, fat droplets are mainly composed by triglycerides for the enhanced accumulation and storage of fatty acids. The fatty acids pool can be increased by three main mechanisms: (a) lipolysis of visceral fat by intracellular lipases; (b) "de novo" liver lipid synthesis; and (c) dietary intake. While fatty liver is an essential marker of NAFLD and ALD, accumulation of lipids in hepatocytes is thought to occur with different mechanisms in these two conditions. In NAFLD patients, using a multiple stable-isotope approach, it was shown that the majority of fat liver deposition (nearly 60%) is coming from increased lipolysis of adipose tissue, whereas dietary fats contribute for only 10%. 14 In this setting, output of fat from adipose tissue F I G U R E 1 Spectrum of liver diseases associated with NAFLD or ALD with the corresponding prevalence. Different pathologic affections encountered in NAFLD and ALD are reported in figure for comparison. Circle area is proportional to the prevalence of any specific condition. In the legend, prevalence is reported in percentage with the corresponding disease. Data were obtained from references. 6

TA B L E 1 Comparison between NASH
and ASH with regard to the mechanisms of liver injury has been demonstrated to be increased by the obese state and proportionate to insulin resistance. 15 Even if increased lipolysis of adipose tissue has been suggested to be involved in ALD (in addition to NAFLD), 16 during chronic ethanol consumption there is also a major contribution of alcohol oxidative processes in increasing nicotinamide adenine dinucleotide (NADH) that in turn stimulates increased fatty acid synthesis 17 and inhibition of fatty acid oxidation. 18 Whatever the specific molecular mechanisms of the onset are, fatty liver has been considered for both NAFLD and ALD the first step of a pathogenetic process possibly leading to liver inflammation/damage.

| P OSS IB LE FAC TOR S CONTRIBUTING TO PROG RE SS I ON FROM FAT T Y LIVER TO NA S H OR A S H
Progression from fatty liver to hepatic injury is coordinately regulated by a series of molecular events. In this perspective, the original "two hits" pathogenetic theory on the development of steatohepatitis 19 has been proposed by the evidence that, more likely, several "parallel hits" contribute to liver damage evolution in NAFLD and ALD. 17,20 On the other hand, since steatohepatitis is a benign condition in a large number of patients, the possibility that NASH may be a different disease rather than a fatty liver evolution is not completely excluded. In the following sub-paragraphs, the possible determinants of damage progression in NAFLD and ALD are described.

| Lipotoxicity
With regard to fats, the increased injurious concentration of lipids and lipid derivatives in hepatic cells has been recognized to determine the so called "lipotoxicity". 21 In the case of NASH, increased liver input of free fatty acids (FFA) is able to determine hepatocyte apoptosis by both intrinsic and extrinsic pathways, evidenced by increased caspase and c-Jun N-terminal kinase (JNK) pathway activity. 22 Also, the production of toxic intermediates (such as diacylglycerol, ceramide and sphingolipids) by dysregulated lipid metabolism further increases liver damage. With regard to lipids, another typical feature of NASH is represented by the accumulation of free cholesterol (FC) within the liver as observed in human tissue by lipidomic analysis. 23 This event has been related to enhanced sterol regulatory element-binding protein 2 (SREBP-2) activity, likely triggered by endoplasmic reticulum (ER) stress. 24 FXR agonists such as obeticholic acid have been shown to improve all parameters of early-stage NASH, effects that were associated with increased LDL-cholesterol. 25 The mechanism by which FC accumulation may promote liver injury and inflammation is not completely clarified yet; however, stimulation of Kupffer cells, ER stress and mitochondrial dysfunction by FC has been observed in experimental models. 26 While lipotoxicity is likely to occur in the course of ASH, this mechanism of injury was not extensively examined in this disease. Indirect evidence, coming from experimental studies, suggests that FA (in particular the unsaturated form) accumulation in the liver, by itself, may promote steatosis and inflammation. 27 However, whether FA playing a role in inflammation processes during ALD remains to be assessed as suggested by a recent review.

| Mitochondrial dysfunction and oxidative stress
Increased fatty content in the liver, during NAFLD, enhances oxidative compensatory events in the mitochondria. 28 As the lipid accumu- and nuclear factor erythroid 2-related factor 2 (Nrf2). 30 These findings identify that DNA in mitochondria is more prone to oxidative damage since this sub-cellular structure does not hold histone or DNA regenerative properties. The progressive cascade of events finally determines a condition defined as mitochondrial dysfunction. 31 Cardiolipin oxidation has also been suggested to have an important role in this process during NASH progression. 32 In fact, a functional cardiolipin is thought to stabilize the structure and function of complex respiratory mitochondrial enzymes, being largely present (more than 20% of total lipids) in the inner mitochondrial membrane. 33 However, this phospholipid given its chemical structure, characterized by several unsaturated groups, is highly prone to oxidative of mitochondrial damage in these two diseases. 35

| Innate immune system in the liver and inflammation
Several innate immune cells are present in the liver, including and endogenous danger-associated molecular patterns (DAMPs). 36 Currently, 10 TLRs have been identified in humans. 37 Among them, TLR4 seems to be the one mainly involved in NASH pathogenesis.

| Background
In recent years, our concept of gut-liver axis has been extended, considering not only the canonical liver (bile-mediated) effects on the gut absorptive process, but also those of gut products (mainly bacterial components and nutrients) reaching the liver through the portal vein. In fact, the liver is the first and more exposed organ to gut-derived products. The gut contains 10-100 trillion of microorganisms  51 This study shows that a change in microbiota composition increases lipid serum levels, which could lead to liver deposition of fats and evolution to steatohepatitis.

| LPS receptors in NASH and ASH
Moreover, increased LPS levels were associated with fed or fasted conditions, while alterations of the gut barrier by increased assumption of ethanol or fats are related to enhanced translocation of microbe-associated molecular patterns (MAMPs) to the liver. 52 Liver interaction with microbial products would result in physiologic or pathologic effects according to different conditions. 53 In this context, TLRs are of paramount importance with their reactivity with PAMPs as described above in regard to IMS. In fact, among PAMPs LPS is the most studied and main activator of TLR4. 54 The latter has a major role for determining inflammatory response in the course of both NASH and ASH, and it has been considered an important biological sensor of circulating LPS since the past century.

| Tight junction integrity and ASH
A study on Caco-2 cells (human colon adenocarcinoma cell line) demonstrated, in a cultured monolayer, that after exposure to ethanol, increased permeability was associated with reduced expression of Zonulin-1 and Claudin-1, two proteins of paramount importance to maintain tight junction integrity. 61 Another study conducted on Occludin knock-out mice showed increased ethanol damage in the IEB of these animals, suggesting a role also for this protein in maintaining a functional non-leaky gut. 62  linking gut-liver axis to liver injury in NASH and ASH is depicted in Figure 3. Recent findings are reported in the following paragraphs.

| Adipokines
Visceral white adipose tissue (WAT), composed by cells deputed to handling of fats (mainly triglycerides and fatty acids), has an important metabolic interplay with the liver. 64 Crosstalk between WAT and liver is mainly supported by an endocrine component of the first, determining the production of several peptide factors collectively named adipokines. 65 WAT dysfunction is suggested in both NASH and ASH since an altered adipokines secretion has been described in these diseases. 66

| microRNAs
microRNAs (miRs), discovered in the early 90s, are small, non-coding fragments of RNA acting at post-transcriptional level and degrading or inhibiting the translation of a specific mRNA. 83  knock-out mice evidenced increased infiltration of fats, inflammation and carcinogenesis in the liver. 86 In agreement with these observations, a study in humans with NASH (n = 25) evidenced a >50% decrease in liver miR-122 expression in comparison with healthy control. 87 Similar results were also observed in mice and humans with alcohol-induced liver damage. 88 In this latter study, in order to investigate the mechanism of miR-122 decrease, the expression of grainyhead-like transcription factor 2 (GRHL2) was evaluated.
GRHL2 was previously found to decrease the expression of this miR in experimental setting of NASH evaluating differentiation of liver progenitor cells. 89 Interestingly, an increased expression (nearly 20fold) of GRHL2 was observed in human liver, with alcoholic cirrhosis.
Moreover, the GRHL2 increase was negatively correlated with miR-122 expression in the same tissue.
On the basis of these results, the pathogenesis of ALD based on GRHL2-mediated miR-122 inhibition has been hypothesized. Other miRs such as miR-34a have been evaluated in experimental models of NASH and ASH. 84  The most relevant miRs altered in NASH and ASH are reported in Table 2. However, possible contribution of miRs altered expression in the pathogenesis of these liver affections remains to be unequivocally demonstrated. Moreover, these post-transcriptional factors seem to act mainly as fine regulators of gene expression, sometimes lack specificity and a phenotypic clear effect, overall suggesting a partial contribution in these multifactorial diseases.

| Extracellular vesicles
It has been demonstrated that a dynamic cell regulation may transpire and TNF-α, suggesting the link between pro-inflammatory cytokines and hepatocyte intoxication during the process of ALD. 95 The comprehensive analysis of liver cell-derived EVs as the therapeutic agents in NASH and ASH will ultimately fill important gaps in our knowledge of the mechanisms of EVs in liver inflammation and gut-liver axis and will address critical barriers to progress in the treatment of the patients with NASH and ASH.

| CON CLUS IONS
We have summarized the translational mechanisms underlying pro-inflammatory signalling and gut-liver axis in non-alcoholic and alcoholic steatohepatitis. NAFLD and ALD represent two important liver diseases in humans, with significantly associated mortality and morbidity. Clinical features are similar suggesting common mechanisms in the onset of these two conditions and possibly a common therapeutic strategy. However, deep examination of proinflammatory processes in NASH and ASH reveals similarities but also important differences. The role of the gut-liver axis is emerging as a major determinant of liver injury in NASH and ASH development and progression. Studies on this aspect will probably help to identify new and novel therapeutic strategies. Characterization of human microbiome remains an important factor to understand these and other liver and non-liver diseases. For this reason, this remains the target of large ongoing, international projects. Another aspect that needs to be examined in deep in the future is that of gender difference in these diseases. Clinical data, for instance, support the view of a protective role of oestrogen against NAFLD since women in premenopausal state or in hormone replacement therapy are seldom affected by this disease. 96 On the other hand, female sex, in humans, seems to increase the risk for toxic liver injury including the ethanol-related one. 97 Some immunological gender differences have been also identified in experimental studies. They may have a possible relationship with NAFLD and ASH damage. As an example, TLR4 is more expressed in normal condition and after stimulation in male human neutrophils. 98 On the other hand, mouse female macrophages have an increased Myd88-related response to LPS. 99 Taken together, these findings recall our attention on the contribution of sex and sex hormones in the pathogenesis of these diseases as in other human affections. Research efforts on these aspects may be helpful to better understand the natural history of these conditions and to identify possible therapeutic targets.

CO N FLI C T S O F I NTE R E S T
The authors confirm that there are no conflicts of interest.

AUTH O R CO NTR I B UTI O N S
TS, LB and FM performed the search and wrote the manuscript; TZ, HF, IL, GG, LK, SL and SG contributed to manuscript writing; and GA and FM supervised the work and wrote the manuscript. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
All the citations and data included in this manuscript are available upon request by contact with the corresponding author.