Gut dysbiosis in nonalcoholic fatty liver disease: pathogenesis, diagnosis, and therapeutic implications

1Nonalcoholic fatty liver disease (NAFLD) is the most common cause of chronic liver injury disease worldwide, with a broad spectrum ranging from simple steatosis, nonalcoholic steatohepatitis (NASH) to cirrhosis and even hepatocellular carcinoma (HCC). Epidemiological investigation shows that NAFLD has become one of the most common liver diseases, with a global prevalence of 25.2%, 27.4% in Asia, but more than 33% in China. As NAFLD is a “silent” disease, the actual number of people with this disease is likely to be higher. Recently, growing evidence displays that NAFLD is multisystem metabolic disease associated with high insulin resistance and genetic susceptibility, in which the accumulation of triglycerides in hepatocytes is the main pathological change. Besides liver-related complications, NAFLD also increase the risk of type 2 diabetes mellitus (T2DM), cardiovascular and chronic kidney disease. Conversely, obesity, T2DM, lifestyle changes and druggable genetic alterations also rise the morbidity and mortality of NAFLD. Despite advance in the pathogenesis of NAFLD, few therapeutic drugs are available. Although nonspecific antioxidative, anti-inflammatory, hypolipemic and hypoglycemic drugs have been widely used in clinic to improve the symptoms, knowledge on NAFLD pathogenesis remains incomplete and hence there are few FDA-approved medications that directly treat NAFLD. In addition, interventions for NAFLD have focused on diet and lifestyle changes, but outcomes are unsatisfactory due to poor patient compliance. Due to the hidden lesions, 20% patients have an irreversible advanced fibrosis and even HCC before the initial diagnosis. Therefore, the early diagnosis is important for ensuring appropriate treatment to delay or control the NASH progression. Current clinical diagnosis of NAFLD involves various procedures such as blood tests, B-ultrasound, transient elastography and MRI, but misdiagnosis of nonspecific symptoms (HBV-associated hepatitis, alcoholic hepatitis, and cirrhosis) is difficult to be distinguished in clinical practice, indicating that effective and reproducible biomarkers are still needed in clinical application.

2The popular theory for the pathological progression of NAFLD is “multi-hit” theory. According to this theory, the vicious cycle of “lipid accumulation/steatosis → hepatic lipotoxicity → metabolic disorders → inflammatory response → insulin resistance → aggravation of metabolic disorders” forms in the liver through interactions among various factors. The dysbiosis of the gut microbiota can be involved in multiple attacks on the liver and plays a key role in the pathogenesis of NAFLD. The intestinal bacteria and their metabolites can enter into the liver through the portal vein and subsequently affect the pathophysiological processes of liver. The gut microbiota and its metabolites can act as the molecular vehicles between intestine and liver. Amazingly, some specific gut microbiota and metabolites, including short-chain fatty acids (SCFAs), bile acids (BAs), lipopolysaccharide (LPS), choline and trimethylamine (TMA), would be changed associated with disease severity and fibrosis stage, indicating the potential of diagnostic markers for NAFLD. Several clinical trials involving gut microbiota intervention, including probiotic and prebiotic supplementation, fecal microbial transplantation (FMT), and microbiome-targeted therapies (MTT), are underway for nearly every type of liver disease. As reported, the overall survival rates of patients with severe liver diseases who received FMT from healthy donors could be increased by 54.2% compared with those without FMT, indicating that the gut microbiota directly affects NAFLD progression. Therefore, the regulation, modification, and reconstruction of the gut microbiota may become a personalized strategy for the prevention and treatment of NAFLD.

3The gut microbiota is a complex community that exist in the human gastrointestinal tract in a symbiotic manner and maintain 3 important physiological functions of the human digestive system, namely digestion, metabolism, and protection. Growing studies have demonstrated several mechanistic pathways of gut microbiota involved in the progression of NAFLD ( Figure 1 ).

4The gut microbiota breaks down undigested polysaccharides into monosaccharides and dietary fibers into SCFAs by utilizing a variety of enzymes, both of which provides energy support for the host cells. Most notably, SCFAs exert multiple benefits in the regulation of metabolism and immune, and could be emerged as a potential therapeutic agent against various diseases. Beside the main source of energy for host colonocytes, SCFAs have distinctive pharmacological and physiological functions, such as promoting colonic motility, protecting intestinal mucosal barrier, altering carbohydrate and lipid metabolism, participating in immune regulation, and improving the absorption of electrolytes and nutrients as well as anti-inflammatory and antitumor activities. Depending on the amount of dietary fiber intake, approximately 500-600 mmol of SCFAs are produced in the colon daily. Among all the metabolic products of carbohydrate fermentation by the microbiota, the acetic acid (C2), propionic acid (C3) and butyric acid (C4) are the most abundant SCFAs in the intestinal tract. Acetic acid is an important source of host energy, providing approximately 10% of total energy daily for the body. Butyric acid as the main source of energy for supplement epithelial cells plays a crucial role in the function of the intestinal barrier; and is the main metabolic substrate for the gastrointestinal microbiota, providing at least 60-70% of its energy requirements for their proliferation and differentiation. It has been demonstrated that the intestinal epithelial cells both in germ-free mice and in antibiotic-induced microbiome depletion mice present shorter turnover and lower proliferative activity than the normal mice, and decreased availability of SCFAs shifts colonocyte energy metabolism toward glucose utilization. Moreover, butyric acid inhibited the activation of ChREBP and SREBP-1, resulting in the repressing of lipogenesis. Propionic acid is mainly catabolized in the liver, where it involves in the transformation from pyruvate to glucose, and it also has been found to reduce lipid accumulation in overweight and adipose sufferers. Unlike butyric acid and acetic acid acting as sources of energy for host cells, propionic acid is the precursor for adipogenesis and gluconeogenesis, which plays more important roles in the pathogenesis of NAFLD. Interestingly, no matter what type of dietary fiber the patients have taken, it will present obvious benefits as soon as it enters the intestinal tract, especially, to those who are already low in dietary fiber intake. Mechanically, SCFAs activate G protein-coupled receptors, such as GPR41 and GPR43, in intestinal and adipose tissues. The activation of GPR41 increases glucagon-like peptide 1 and peptide YY (PYY) secretion by enteroendocrine cells, resulting in the decreased intestinal motility but enhanced nutrient absorption. GPR43 activation inhibits adipocyte differentiation, and increases hepatic lipogenesis, thereby promoting the development of NAFLD.

5BAs are the main components of bile and can be divided into primary and secondary bile acids. Primary bile acids (PBAs), including cholic acid (CA) and chenodeoxycholic acid (CDCA), are produced from cholesterol in hepatocytes and excreted to the bile duct. Secondary bile acids (SBAs), including lithocholic acid (LCA), deoxycholic acid (DCA), ursodeoxycholic acid (UDCA) and their respective isoforms (such as isolithocholic acid), are then transformed from the primary bile acids by bacteria in the small intestine. PBAs not only maintain intestinal microbiota homeostasis through directly inhibiting overgrowth of pathogenic bacteria, but also act as the naturally occurring agonists of intestinal FXR (FXR activation: CDCA>DCA>LCA>CA; FXR inhibition: UDCA) and subsequently activate its downstream defense genes in ileal mucosa to protect intestinal epithelial cells against bacterium and microorganism’s corrosion. In addition, FXR down-regulates the expression of LXR and SREBP-1c to reduce fatty acid and triglyceride synthesis in the liver, thereby reducing steatogenesis and gluconeogenesis. FXR also up-regulates hepatic glycogen synthesis via the activation of fibroblast growth factor (FGF) 15/19, PPARγ, GLUT-4 and GLP-1 to improve insulin sensitivity. TGR5 is another classical BA receptor, mainly activated by SBAs (TGR5 activation: LCA>DCA> CDCA>CA). Too few SBAs lead to reduced FXR activity but increased inflammation in the body, while too many SBAs can cause cellular DNA damage through the production of reactive oxygen species (ROS), leading to the development of HCC. It had been found that activation of TGR5 by SBAs induces transcription of the type 2 iodothyronine deiodinase (Dio2) gene, which in turn converts thyroid hormone (T4) to the more active triiodothyronine (T3), thereby increasing basal metabolism and promoting energy metabolism in brown adiposes and muscle tissues of the high-fat diet (HFD)-fed mice. Activation of TGR5 in intestinal secretion cells by SBAs promotes the expression of GLP1, which increases insulin synthesis and release, resulting in the protection against islet β-cell apoptosis and improvement of blood glucose. However, alteration of gut microbiota and hepatic function directly affects the ultimate compositions and the contents of BAs, resulting in the differential expression profiles of BA receptors (e.g., FXR vs. TGR5) among different stages of NAFLD process. Anyway, these two BA receptors, FXR and TGR5, have emerged as putative therapeutic targets for obesity and NAFLD.

6Choline and its derivatives, such as TMA and TMAO, are the major metabolites of intestinal microorganisms, and their elevated levels not only contribute to atherosclerosis but also are closely related to cholesterol and triglyceride metabolism. Clinical trials have found that reduced choline can lead to lipid deposition in the liver by reducing the synthesis and secretion of very low-density lipoproteins (VLDL) in hepatocytes, leading to the steatohepatitis. It is the consequence which could be commonly found in methionine/choline-deficient (MCD) diet-fed rodents. It has been known that the food containing dietary methylamine, choline, phosphatidylcholine and carnitine will break down into various metabolites including TMA through the catabolism of trimethylamine lytic enzymes in Proteobacteria and Firmicutes. TMA is transferred to the liver via the portal vein and converted by flavin-containing monooxygenases into TMAO, which promotes the accumulation of activated leukocytes into human endothelial cells, leading to endothelial dysfunction and significantly increases the risk of atherosclerosis and cardiovascular diseases.

7Growing investigations held the intestinal barrier dysfunction blame for inflaming the NAFLD progression. During the initiation and progression of NAFLD, large amounts of gut bacterial metabolites, bacterial components, and other hazards enter the liver through the portal vein due to the intestinal mucosal barrier disruption induced by various stimuli, with the increased intestinal permeability. Those attacks can further accelerate liver injury and fibrosis via elevating inflammation, oxidative stress, and lipid accumulation as shown in Figure 2. The investigation by in situ hybridization revealed the presence of gut bacterial metabolites and DNA fragments in the livers of HFD-fed mice, but presence of bacteria in the liver of NASH patients remains unkown. Compared with normal individuals, the number of enteric bacteria, especially Gram-negative bacteria, was significantly increased in obesity or NAFLD patients, and presented obvious endotoxemia. The excessive amounts of LPS activates adenylate cyclase in the intestinal mucosa and damages the mitochondria and lysosomes of epithelial cells, leading to necrosis of the apical cells of the intestinal villi and autolysis of the epithelial cells. Moreover, the gut-derived lipopolysaccharides (LPS) are the key factor in conducting the liver inflammation and chronic damage through activating LPS-dependent pattern recognition receptor signaling. LPS activates TLR4 in endothelial cells and TLR9 in dendritic cells, resulting in the release of a large number of pro-inflammatory cytokines (such as TNF-α, IL-1β, and IL-6) and chemokines (CCL2, CXCL2, CXCL10 and CXCL16) that mediates inflammation and pathological damage to the liver. Therefore, LPS causes inflammation and serious metabolic changes in the body, such as increasing fat consumption, and elevating circulating free fatty acid (FFA) and triglyceride (TG). The deposition of FFA in the liver may also induce inflammation, and insulin resistance (IR), further contributing to the development of NAFLD.

8On the other hand, the gut microbiota can alter the balance of anti-inflammatory and proinflammatory cytokines secreted by M1 and M2 macrophages by affecting the metabolism of SBAs, thereby affecting the immune function of the liver. Even small amounts of SBAs produced by the gut microbiota can decrease FXR activity and increase inflammation in the body; however, high amounts of SBAs can produce large amounts of ROS, cause cellular DNA damage, and lead to the development of HCC. In addition, PBAs can upregulate CXCL16 in hepatic vascular endothelial cells, which in turn leads to the recruitment of NKT cells that can kill tumor cells in a CD1d-dependent manner.

9Several studies have shown that Patients with NAFLD have the remarkable gut dysbiosis, in which the relative abundances of Proteus and Enterobacter bacteria were increased while and Ruminococcus and Lactobacillus were decreased. As simple steatosis progresses to advanced liver fibrosis, the number of Gram-negative bacteria was increases, especially Proteus bacteria. Fecal BAs may reflect alterations in two bile acid metabolic pathways: “glycine, serine, and threonine metabolism” and “taurine and hypotaurine metabolism.” Escherichia, Bilophila and Rhodobacter involved in BAs-mediating taurine and glycine metabolism were significantly increased in the faeces of NASH patients. Moreover, 60% of NAFLD patients were associated with infection with high alcohol Klebsiella pneumoniae (HiAlc-Kpn), and more interestingly, transplantation of this HiAlc-Kpn isolated from the feces of NASH patients could induce NAFLD in the mice. However, the profiles of gut microbiota found in different studies often vary greatly due to differences in study subjects, dietary habits, medication history, and detection methods. Although close interactions between gut microbiota and NAFLD pathogenesis have been demonstrated over the past decades, the reliable microbiome markers for early NAFLD still remain to be further verified. Recently, with the aid of machine learning model integrating baseline microbial signatures to distinguish liver fat accumulation status of NAFLD participants has displayed good perspective in clinical practice, with 80% of overall accuracy (vs 60% by blood biochemical detection). Given the obvious individual differences and dynamic changes of gut microbiota, it is difficult to classify disease status by stool-based biomarker detection using 16s rDNA amplification alone rather than multi-omics analysis, such as metagenomics and other culture-independent technologies. To increase the robustness and generalizability of predictors, the set of predictive metabolites identified by multiple methods could be more potential for future trials to validate prediction biomarkers in larger and independent cohorts.

10Up to now, the basic steps in the prevention of NAFLD are still diet, exercise and management of complications (such as type 2 diabetes and dyslipidemia). There are few effective drugs for treatment and prevention of NAFLD/NASH. Pioglitazone may be effective in advanced NASH patients or associated with type 2 diabetes, but reliable clinical data are lacking; vitamin E and D have some efficacy, but the safety of long-term use is unclear and the course of treatment is inconclusive; statins can reduce serum LDL levels and prevent cardiovascular complications, but do not address the progression of liver disease. Some promising drugs, including BA derivatives, BA metabolic-mediated agents, and probiotics, are still being evaluated in clinical trials ( Table 1 ).

11MTT is a novel strategy for the treatment of NAFLD by correcting or reshaping gut microbiota by using antibiotics, probiotics, prebiotics, synbiotics and FMT to maintain gut homeostasis. Commercialized Streptococcus, Lactobacillus, and Bifidobacterium improve the gut inflammatory microenvironment, promote the growth and survival of intestinal epithelial cells, and inhibit pathogenic bacteria by modulating the immune system and host defense. As safe edible probiotics, Lactobacillus and Bifidobacterium can lower blood cholesterol levels. FMT is a comprehensive approach for regulating the gut microbiota that has been shown to be effective in restoring microbiota composition and can be used to treat C. difficile-associated diarrhea and Pseudomembranous colitis, as well as other chronic intestinal inflammatory diseases, such as ulcerative colitis and Crohn’s disease. After FMT treatment, the gut microbiota in both HFD-fed mice and NAFLD patients would be reshaped, small intestinal permeability was elevated, and the symptom of NASH was partly improved. However, unlike the necessity for ulcerative colitis and Crohn’s disease, FMT is not often the first choice for NAFLD patients due to the distaste for the gut contents. Compared with FMT, prebiotic supplementation has a higher safety and operability. The oral administration of prebiotics, such as lactulose, oligofructose, and inulin, can stimulate the release of gastrointestinal peptides, which have the effect of regulating appetite and energy metabolism of the body. However, the excessive intake of inulin may cause liver damage, and thus prebiotics should be supplemented appropriately under the guidance of a doctor to prevent harm to the body. Although probiotics, prebiotics and synbiotics have been proposed for treating the patients with obesity-related NAFLD and indeed alleviate intestinal permeability, they did not show any significant clinical improvement in NAFLD patients. Therefore, their therapeutic effects should be further investigated by high-quality clinical trials.

12The use of a specific bacteria or bacterial products to interfere with NASH may be a novel therapeutic strategy for anti-NAFLD drug discovery. The colonization of Akkermansia muciniphila in gut could improve liver function, reduce oxidative stress, inhibit inflammation, normalize gut microbiota, and reverse metabolic disturbances caused by HFD. Some SBA-producing bacteria, such as Lactobacillaceae and Lachnospiraceae, exhibit cholesterol-lowering potential and may improve metabolic abnormalities under HFD.

13Recently, phage therapy is being given great attention as a new antimicrobial agent. Understanding endogenous phage-gut microbiota interactions in health and in NAFLD may enable phage utilization in precise gut microbiome editing, towards treating NAFLD and other obesity-associated metabolic diseases. As previously reported, the targeted eradication of HiAlc-Kpn with phages effectively alleviates bacterial auto-wine syndrome and NASH in NAFLD model mice. Treatment with bacteriophages can specifically target cytolytic Enterococcus faecalis and alleviate liver injury in humanized mice colonized with faecal bacteria of patients with alcoholic hepatitis. Unlike the major therapeutic hurdles of antibiotics and other chemotherapies caused by high degrees of bacteria resistance in clinic, phage therapy provides a new therapeutic approach for precisely editing the intestinal microbiota, completely resolving bacteria resistance and treating liver diseases. Although great advances have presented in various bacterial infection, due to the diversity and complexity of gut microbiome, more clinical trials with larger cohorts are required to validate the relevance of the findings in NAFLD patients.

14The gut microbiota also has been demonstrated to enhance the therapeutic response to immune checkpoint inhibitors though targeting immunomodulatory molecules or their ligands on the surfaces of T cells. The combination has been approved for treating various malignant tumors, including NAFLD-associated HCC. Compared with those untreated with antibiotics, progression-free survival and overall survival of patients treated with antibiotics were significantly lower during the treatment of anti-PD-1/PD-L1 antibodies. This finding was consistent with the results of the mouse experiment, in which the transplantation of fecal flora from tumor patients who responded to PD-1 inhibitor into germ-free mice enhanced the antitumor effects of PD-1 inhibitor, while Fecal flora from non-responding patients did not exhibit this improvement. These results suggest that the regulation of gut microbiota may be a new target for NAFLD-related HCC and cirrhosis therapy. Given that patients with HCC and cirrhosis often have severe intestinal ecological dysregulation, which may lead to immunotherapy failure in some patients, the modulation of the gut microbiota may be more relevant for the therapeutic response in HCC and cirrhosis than other solid tumors.

15Notably, recent data indicate that the IL-17/IL-17R originating from TH17 cells drives intestinal neutrophil migration, limits gut dysbiosis, maintains intestinal barrier integrity and attenuates LPS translocation to visceral adipose tissue, resulting in protection to metabolic syndrome and comorbidities such as NAFLD and obesity. Paradoxically, there is also some evidence that IL-17 accelerated hepatic steatosis through activation of the JNK-PPARα pathway in the HFD mice and oleic acid-preloaded hepatocytes. Circulating levels of IL-17, released by the visceral adipose tissue, induces eotaxin secretion via the smooth muscle cells, both of which are associated with early atherosclerosis that is the main factor drastically reducing the survival of NAFLD patients. These results indicated the versatile of IL-17 in the development of NAFLD and the significance of target delivery in different tissues.

16Besides the central roles in bile acid metabolism, FXR and TGR5 are closely linked to the metabolism of lipids, glucose and lipoproteins, and hence have emerged as the new targets for the treatment of NAFLD. FXR agonists include natural ligands (such as CDCA), semi-synthetic BAs (such as obeticholic acid (OCA)) or synthetic non-steroidal molecules (such as GW4064, WAY-362450). These FXR agonists prevent hepatic steatosis in obese and insulin-resistant rodents. Synthetic FXR agonists are associated with increased body mass and reduced glucose tolerance with prolonged use, whereas the BA class of natural FXR agonists are not. There are two published clinical trials on the FXR agonist OCA. Short-term oral administration of OCA improved insulin resistance and reduced markers of liver fibrosis in patients with NASH combined with type 2 diabetes. However, in the FLINT (clinical trials gov identifier NCT01265498) trial, long-term oral administration of OCA at 25 mg/d for 72 weeks in patients with NASH did not improve insulin sensitivity and increased homeostasis model assessment (HOMA) indices. The difference between these two trials may lie in the length of time the medication is taken. Indeed, OCA treatment significantly improved liver histological outcomes, with a reduction in NASH activity scores compared with placebo-treated patients. However, the improvement rate was only 22% and hardly seen in patients with severe fibrosis. Also, the adverse reactions caused by FXR agonist are the key problems that have to be faced. Therefore, further clinical trials are needed to assess the efficacy and safety of OCA for the clinical treatment of NASH.

17Up to now, most synthetic FXR agonists are ongoing at preclinical or phase I status. Recent study has shown that hepatic FXR subtypes (α1 and α2) differ in their mechanisms of limiting lipid accumulation in hepatocytes, in which FXRα2 has a stronger inhibition on hepatic TG levels. This has implications for the further optimization of FXR agonists and the improvement of therapeutic efficacy.

18TGR5 agonists are theoretically able to maintain glucose homeostasis, increase energy consumption to maintain healthy body mass, and reduce liver inflammation in NASH. The CA derivative INT-777 is a selective TGR5 agonist that increases energy metabolism, reduces liver fat deposition and prevents body mass gain in HFD-fed mice. The TGR5/FXR dual agonist INT-767 showed good affinity to both receptors, and reduced hepatic steatosis in obese diabetic mice via repressing the production of pro-inflammatory cytokines and inducing M2 macrophage polarization. Despite good results in animal studies, clinical trials with TGR5 agonists have been less effective.

19UDCA is a classical drug with immunomodulatory, anti-oxidant and anti-apoptotic effects. However, its effect on NAFLD/NASH has not been confirmed in clinical trials. However, some derivatives of UDCA have shown good therapeutic effects in mouse NASH models, such as norUDCA, which is currently in phase II clinical trials.

20The BA sequestrants colesevelam interrupts the hepatic-intestinal circulation of BA and has been shown to affect cholesterol, triglyceride and glucose homeostasis, reducing hepatic fat deposition in NASH patients. Several additional ASBT antagonists have been developed for the treatment of hyperlipidemia and hypercholesterolemia, but have not been tested in NAFLD.

21Abnormal gut microbiota can promote the occurrence, progression, and even deterioration and carcinogenesis of NAFLD. How to use gut microbiota to alter or block the progression of NASH and find bacteria and bacterial products, as potential targets, has become the critical for NAFLD intervention and treatment. First, although new strategies, such as probiotics, prebiotics, FMT, phage therapy and FXR/TGR5 agonists, have shown promise in the treatment of NAFLD, challenges in developing safe and effective therapeutic drugs, including the selectivity, tissue specificity and drug resistance for long-term use, remain further assessed.

22Second, whether specific bacterial strain or microbiota are causally related to the occurrence of NASH and could be served as effective therapeutics has not yet been established mechanistically. Moreover, most studies have remained at the animal stage, and how differences in gut bacterial species between animals and humans affect microbial metabolism remains unclear.

23Third, other microbial factors, such as metabolites associated with gut virus, may be also important to the progression of NAFLD. To date, the microbial origins of many detected metabolites are unclear, and few microbial sources have been reported for several metabolites. The metabolomic profiles of patients with NAFLD are inconsistent and sometimes contradictory probably because of the relatively small number of patients and the lack of uniform methods and standards. Therefore, in-depth studies are required to explore these metabolites and their microbial sources and to refine the characteristics of floras and their metabolites in different NAFLD subtypes. It is believed that the developments in technology and research may facilitate the use of the gut microbiota in NAFLD intervention or treatment, and the in-depth exploration of the composition and function of gut microbiota will bring new diagnostic tools and personalized treatments for NAFLD.

24JF, and C-HY: Wrote the manuscript; X-JL, J-MY, S-HY, and Z-YF: Searched literatures; C-HY, and JF: Illustrated figures and tables; Z-YF, and W-YY: Edited the manuscript; C-HY and W-YY: Designed and supervised the review. All authors contributed to the article and approved the submitted version.

25This work was supported by Natural Science Foundation of Zhejiang Province (No. LZ17H280001 and LY19H200003), Key research projects of Hangzhou Medical College (No. KYZD202106), and Key Laboratory of Neuropsychiatric Drug Research of Zhejiang Province (No.2019E10021).

26The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

27All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.