Pathogenesis of NASH: The Impact of Multiple Pathways

Abstract

Purpose of review

Advancing our understanding of the mechanisms that underlie NASH pathogenesis.

Recent findings

Recent findings on NASH pathogenesis have expanded our understanding of its complexity including: (1) there are multiple parallel hits that lead to NASH; (2) the microbiota play an important role in pathogenesis, with bacterial species recently shown to accurately differentiate between NAFL and NASH patients; (3) the main drivers of liver cell injury are lipotoxicity caused by free fatty acids (FFAs) and their derivatives combined with mitochondrial dysfunction; (4) decreased endoplasmic reticulum (ER) efficiency with increased demand for protein synthesis/folding/repair results in ER stress, protracted unfolded protein response, and apoptosis; (5) upregulated proteins involved in multiple pathways including JNK, CHOP, PERK, BH3-only proteins, and caspases result in mitochondrial dysfunction and apoptosis; and (6) subtypes of NASH in which these pathophysiological pathways vary may require patient subtype identification to choose effective therapy.

Summary

Recent pathogenesis studies may lead to important therapeutic advances, already seen in patients treated with ACC, ASK1 and SCD1 inhibitors and FXR agonists. Further advancing our understanding of mechanisms underlying NASH pathogenesis and the complex interplay between them will be crucial for developing effective therapies.

Introduction

With current data showing that nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease, with an estimated worldwide prevalence of 25.24% [1], advancing our understanding of the pathogenesis of the disease is crucial. Approximately 25–40% of NAFLD patients will progress from the more benign form, nonalcoholic fatty liver (NAFL), to the more advanced form of the disease, nonalcoholic steatohepatitis (NASH) [2–4], with hepatic steatosis, lobular inflammation, ballooning, and/or perisinusoidal fibrosis that may ultimately lead to hepatocellular carcinoma or cirrhosis and liver decompensation. For the latter, liver transplant is the only current treatment option, with the most recent U.S. data showing that NASH is the most common cause of cirrhosis [5] and the current leading indication for liver transplant in women and the second leading cause for men [6]. The mechanisms that underlie the pathogenesis of the disease are not yet fully elucidated but recent advances in our knowledge of these may soon lead the way toward the development of more effective treatments.

Although it was once proposed that NASH was the result of two sequential “hits” (pathways), the most recent understanding is that there are multiple parallel hits that result in the inflammation, cell death, and fibrosis seen with the disease. Included are oxidative stress, endoplasmic reticulum (ER) stress, altered lipid metabolism, altered production of adipokines and cytokines, mitochondrial dysfunction, lipotoxicity, glucotoxicity, gut-derived endotoxin, and genetic predisposition [7–9]. With the continuing identification of additional factors that contribute to the development of NASH and the many ways in which these factors interact its pathogenesis is now understood to be extremely complex. We discuss in this review the most well accepted and recent findings.

Factors that Underlie Predisposition to NAFLD

A number of factors, including some only very recently identified, are known to predispose the development of NAFLD and NASH (Figure 1).

Figure 1:

Pathogenesis of NAFLD/NASH and Contributing Factors

Genetics

Genome-wide association studies have identified dozens of genes with multiple polymorphisms that are associated with fatty liver risk in specific populations [10]. However, only two genes, PNPLA3 and TM6SF2, have been consistently identified as modifiers of risk. The rs738409[G] allele of PNPLA3, encoding Ile148Met, has been consistently shown to be associated with higher liver fat content, higher necroinflammatory scores, and a substantially increased risk of developing fibrosis [10]. However, it should be noted that the association between this variant and NAFLD is not linked to metabolic syndrome, with this PNPLA3 allele driving disease development in combination with other factors such as lifestyle, viral infection, overconsumption of alcohol or other cryptogenic causes [10]. A recent study showed that the I148M PNPLA3 variant disrupts the ubiquitylation and proteasomal degradation of PNPLA3 which leads to accumulation of PNPLA3-I148M and impairment in the mobilization of TGs from lipid droplets [11]. Another recent report has shown that PNPLA3-I148M potentiates the profibrogenic and pro-inflammatory features of hepatic stellate cells (HSCs) [12]. The other genetic variant that has been widely validated as being associated with NASH is the rs58542926 (c.449 C > T) allele of TM6SF2 which encodes an E167K amino acid substitution. It seems clear that the TM6SF2 E167K variant is associated with increased risk for progressive NASH although, interestingly, a recent study shows that the variant may also be associated with reduced risk of cardiovascular disease [13]. Many other genes involved in carbohydrate and lipid metabolism, insulin signaling pathways, inflammatory pathways, oxidative stress and fibrogenesis have been shown to play a role in NAFLD/ NASH. Some of those include GCKR, APOB. LPIN1, UCP2, IFLN4, the newly discovered variant HSD17B13 and others [14–19]. The HSD17B13 is also a lipid trafficking protein present on lipid droplets and confers protection from liver disease. A splice variant is associated with increased risk of NASH. The fact that key lipid trafficking proteins are related to the risk of NASH indicate that lipid trafficking plays a major role in disease pathogenesis. This relationship requires further elucidation.

Epigenetics and microRNAs

Multiple epigenetic aberrations have also been associated with pathogenesis. These epigenetic changes have been shown to be associated with hepatic lipid metabolism regulation, insulin resistance, mitochondrial dysfunction, oxidative stress, ER stress and the release of inflammatory cytokines [20]. Epigenetic modifications usually occur through DNA methylation, protein acetylation and/or micro RNAs (miRNAs). An epigenetic study in humans has shown that some methylated genes (FGFR2, MAT1A and CASP1) can distinguish between patients with advanced NASH and those with simple steatosis [21]. MAT1A is responsible for S-adenosylmethionine metabolism and is part of the glutathione cycle, all of which may play a role in NAFLD and NASH [22].

The liver expression of certain miRNAs, including miR-181a, miR-34a, miR-122, miR-200 and miR-192, has been shown to correlate with the histological features of NASH [23]. More studies are needed to explore their mechanisms but some of those have recently been discovered. Examples are the roles of miR-141/200c in diminishing NASH-associated hepatic steatosis and inflammation through reprogramming of lipids and inflammation signaling pathways [24] and of miRNA-21 in decreasing inflammation and fibrosis via the restoration of PPARα expression [25].

Systemic Milieu in Which NASH Develops

Diet

Caloric intake and nutrient composition play a key role in NAFLD (Figure 1). Fructose intake is associated with hepatic steatosis, obesity and insulin resistance [26]. It plays a key role in triggering hepatic inflammation and subsequently in developing NASH. Saturated fat induces de novo lipogenesis, ER stress and apoptosis [27]. Trans fat intake is associated with NAFLD [28]. Cholesterol, iron overload and low copper are all associated with NASH [29–31]. The Western diet includes high amounts of saturated fat and omega-6 (n-6) polyunsaturated fatty acids (PUFAs) and low amounts of omega-3 (n-3) PUFAs [27]. This imbalance has been shown to be associated with inflammation and NASH development [27]. A recent study showed that red meat and processed meat are associated with insulin resistance and NAFLD [32]; larger studies are needed to confirm this finding.

Adipose tissue and adipokines

The adipose tissue plays a critical role in NAFLD progression through the release of adipokines, including adiponectin and leptin, and cytokines, including TNF-α and IL-6. Once the adipose tissue mass is increased the balance between adipokines and cytokines is lost leading to insulin resistance, obesity and hepatic steatosis. Leptin is mainly founded in the adipose tissue and is important for energy homeostasis and neuroendocrine function (including, for example, appetite). Increased levels of leptin initially lead to inhibition of both hepatic glucose production and de novo lipogenesis via stimulation of fatty acid oxidation, but as the pathological process continues, leptin may promote inflammation and fibrogenesis [33]. In the liver, leptin acts through its receptor, leptin receptor type b (LepRb), to decrease the expression of sterol regulatory element-binding transcription factor 1 (SREBP-1) which regulates glucose, fatty acid and lipid metabolism. Leptin also upregulates transforming growth factor type 1 (TGF-1) which in turn stimulates stellate cells giving leptin fibrogenic activity [34]. In a recent meta-analysis it was found that both patients with NAFL and NASH have higher leptin concentrations [35]. It was also recently shown that leptin is not associated with response to treatment in NASH patients treated with vitamin E [36]. However, there is evidence from lipodystrophy patients with NAFLD/NASH that leptin treatment may have beneficial effects including histological ones [37–40].

Adiponectin, one of the most abundant adipokines, is produced by the liver in response to liver injury. It demonstrates antisteatotic and antiapoptotic effects on hepatocytes and exerts anti-inflammatory and antifibrotic effects acting on Kupffer cells and HSCs [41]. Once the adipose mass is increased adiponectin levels decrease due to multiple factors including oxidative stress and mitochondrial dysfunction [42, 43]. Interestingly, once NASH progresses toward advanced fibrosis, circulating adiponectin increases, possibly due to decreased hepatic clearance and/or a compensatory mechanism to increase the release of inflammatory cytokines [44]. This phenomenon is one of the reasons why steatosis decreases in advanced stages of fibrosis in NASH [45].

Pro-inflammatory state and changes in microbiome

It has been shown that there is a leaky gut in multiple liver diseases, including NASH, that is associated with translocation of lipopolysaccharides (LPS) through the portal circulation to the liver, predisposing to NASH [46]. Circulating endotoxins have been found to be elevated in NASH patients, most likely related to the gut leak phenomenon seen during the disease process [47, 48]. Recent research has shown that certain chemokines play a role in regulating gut permeability and may play a role in NASH development [49]. Multiple recent studies have also provided substantial evidence that the gut microbiota play an important role in the pathogenesis of NASH, with a state of dysbiosis clearly present in NASH patients [47, 50].

In a recent study, bacterial species could differentiate between NAFL and NASH patients with good accuracy [51]. However, studies that have assessed specific microbiota alterations in NASH have had inconsistent and sometimes flatly contradictory findings. For example, the findings of three recent studies that compared NASH patients to healthy controls showed a reduced presence of Bacteroidetes [52], a higher abundance of Bacteroidetes [53], and no difference in Bacteroidetes levels but a decrease in Firmicutes [54]. A number of mechanisms may underlie the microbiota’s effects on NASH, including alterations in gut permeability, possibly linked to dysregulation of epithelial tight junction formation; defective inflammasome sensing and disrupted inflammatory responses; altered choline metabolism by the microbiota; bacterial metabolites produced, degraded, or modulated in the gut including LPS, short chain fatty acids (including acetate, propionate, and butyrate) whose production and absorption are increased, and bile acids; and increased delivery to the liver of ethanol produced by the gut microbiota [50, 10].

Crosstalk of Multiple Pathways Leading to Hepatic Injury and NASH

Hepatic steatosis and lipotoxicity

The progression toward fatty liver begins with hepatic accumulation of lipid droplets in the cytoplasm of hepatocytes. This results from imbalanced metabolism of lipids so that the influx and synthesis of lipids exceeds the capacity of the hepatocyte to secrete or utilize them. There are three major sources of FFAs in the liver including uptake of FFAs from adipose tissue, diet and de novo lipogenesis (Figure 2). In absolute terms, FFAs derived from adipose tissue lipolysis are the largest source of FFAs in the liver whereas DNL is increased and is relatively the most altered compared to normal individuals. Adipose tissue lipolysis, the key mechanism of adipose tissue-derived FFAs, is a direct function of insulin sensitivity and is increased in insulin resistant states. This results when (1) increased free fatty acid influx from adipose tissue to the liver, (2) increased fat delivery from the gut and adipose tissue to the liver, and (3) increased liver de novo lipogenesis are not sufficiently countered by increased very-low-density lipoprotein (VLDL) secretion or use of lipids in other pathways [8, 55].

Figure 2:

Role of Free Fatty Acids in NAFLD/NASH (Lipotoxicity and Glucotoxicity)

Abbreviations: FFA: Free Fatty Acid, VLDL: Very low density lipoprotein

Multiple types of lipids are known to be involved in the development of hepatic lipotoxicity, the effect on liver cells that results from elevated concentrations of lipids and their derivatives [8]. Among the lipids and their derivatives whose levels increase are free fatty acids (FFAs), triglycerides (TGs), free cholesterol, lysophosphatidyl choline (LPC), bile acids, and ceramides. However, increases in these are not all harmful. For example, increased TG levels appear to be protective against excessive accumulation of toxic TG precursors [9, 56].

The main drivers of liver cell injury appear to be the lipotoxicity caused by FFAs and their derivatives combined with mitochondrial dysfunction [56, 57]. The saturated fatty acid palmitic acid has been shown in cell studies to activate peroxisome proliferator-activated receptor alpha (PPAR-α), impair insulin signaling, and cause c-Jun N-terminal kinase (JNK)-dependent mitochondrial dysfunction and caspase activation that result in cell death [8]. Palmitic acid also increases the level of intracellular LPC which itself causes ER stress and induces apoptosis.

In NASH, elevated levels of ceramides correlate with elevated levels of pro-inflammatory cytokines, including IL-1, and IL-6, and are associated with hepatic insulin resistance, oxidative stress and inflammation. It has been shown that the interaction between tumor necrosis factor-α (TNF-α) and ceramides results in the production of reactive oxygen species (ROS) in the liver cells’ mitochondria, increasing inflammation and resulting in apoptosis.

The combination of increased free cholesterol synthesis and de-esterification with decreased bile acid synthesis and cholesterol export results in an over-abundance of free cholesterol in the liver. This leads to liver damage by activating signaling pathways in hepatocytes, HSCs, and Kupffer cells that promote fibrogenesis and inflammation [58]. In addition, the excess free cholesterol in the liver mitochondria results in mitochondrial dysfunction, with resulting increased levels of ROS which in turn stimulate the unfolded protein response (UPR) in the ER, ultimately leading to ER stress and apoptosis [58].

In NASH, the hydrophobic bile acid deoxycholic acid (DCA) induces JNK1 phosphorylation which appears to induce both p53 expression and activation, converging in a strong and functional engagement of the miR-34a-dependent apoptotic pathway [59]. The increase can induce production of ROS resulting in apoptosis and cell death. Because when they bind to bile acids the nuclear hormone receptors Farnesoid X receptors (FXR) repress the expression of the gene encoding the rate-limiting enzyme in bile acid synthesis, cholesterol 7 alpha-hydroxylase, there is ongoing research assessing the effectiveness of FXR agonists for NASH treatment.

Hepatic glucotoxicity

Hepatic glucotoxicity is the toxic effect on the liver cells and tissues of excess carbohydrate intake and hyperglycemia. The metabolic changes that result from continuing hyperglycemia cause low-grade inflammation which contributes to insulin resistance. Both oxidative stress and ER stress induced by elevated sugar intake are also major contributors to decreased insulin sensitivity [8].

Dietary fructose has been shown to be highly lipogenic in humans [60]. Fructose intake increases lipogenesis, hepatic fat deposition and inflammation through multiple factors including increases in fatty acyl coenzyme A, diacylglycerol (DAG), triacylglycerides (TAG), the fat transporter CD36, and the lipogenic enzymes fatty acid synthase (FAS), SCD-1, carbohydrate responsive element binding protein (ChREBP) and ACC [8]. With increased fructose intake, there is a reduced ability of the body to dispose of glucose and restricted insulin action in the liver. Although increased fat intake alone may result in reduced insulin sensitivity, the addition of high fructose to high fat intake worsens insulin resistance and leads to increased inflammation, fibrosis, ER stress, and lipoapoptosis in the liver [8].

In animals fed high-sugar diets oxidative markers are elevated and antioxidant status is reduced, with decreased levels of reduced glutathione, glutathione peroxidase, and catalase [61–63]. Increased JNK activity may also play an important role in creation of insulin resistance through inhibition of IRS-1 which blocks insulin signaling [64]. Elevated carbohydrate intake is also known to activate multiple lipogenic enzymes that can induce lipogenesis and steatosis, including fatty acid synthase (FAS), acetyl-CoA carboxylase (ACC), and stearoyl-CoA desaturase-1 (SCD-1). Multiple therapeutic agents that suppress ACC and SCD-1 are currently under investigation for NASH treatment [65, 66]. The overabundant carbohydrates can be initially converted into TGs and FFAs after which multiple hepatotoxic lipids accumulate, including LPC, free cholesterol, ceramides, and bile acids.

Cell Stress and Apoptosis

Mitochondrial dysfunction and oxidative stress

It has been shown that in NASH there is elevated mitochondrial mass but 30–40% lower maximal respiration with associated mitochondrial uncoupling and leaking along with increased hepatic insulin resistance [67]. Both genetic factors and epigenetic changes are thought to contribute to mitochondrial dysfunction in NASH [9]. Under physiological conditions fatty acids are oxidized via B-oxidation and then get transported to the mitochondrial respiratory chain (MRC). Long-chain fatty acids are oxidized and transported in the mitochondria. This process is associated with reduction of oxidized NAD+ and FAD to NADH and FADH2. Once FFA flux is increased, the mitochondria become exhausted and FFAs are handled in other sites including peroxisomes (β-oxidation) and the endoplasmic reticulum (ω-oxidation) [68]. Mice fed with a high-fat, high-sucrose diet have been shown to have lower hepatic NAD+ levels leading to impaired hepatic mitochondrial functions, steatosis and lipid peroxidation [69, 70]. Once the mitochondria are exhausted the partially reduced oxygen molecules lead to oxidative stress [68]. The mitochondrial CYP2E1 then tries to oxidize the rest of the excess FFAs which further increases ROS production; CYP2E1 has been shown to be increased in patients with NASH [71, 72]. The remaining excessive FFAs also get oxidized in the peroxisomes in which electrons from FADH2 and NADH lead to further formation of ROS. These processes in the mitochondria cause mitochondrial dysfunction evidenced by decreased mitochondrial DNA levels and proteins and ATP depletion. A new mitochondrial player, Sab (Sh3bp5), an outer membrane mitochondrial protein and binding target of JNK, has been described [73]. In hepatocytes exposed to palmitic acid, SAB and JNK, interplay led to mitochondrial dysfunction and apoptosis. Another carrier complex, mitochondrial pyruvate carrier (MPC), located in the inner mitochondrial membrane, may also play a role in NASH treatment. MPC2 is a mitochondrial binding site for thiazolidinediones; in a recent study in mice it was shown that treatment with a thiazolidinedione led to attenuating fibrosis [74]. In addition to the cross talk between the ER and mitochondria the neighboring lysosomes can also contribute to mitochondrial dysfunction [75, 76]. FFAs can translocate into lysosomes leading to lysosomal enzymes into the cytosol, subsequently leading to activation of nuclear factor (NF)-κB and TNF-α, in turn leading to JNK activation, mitochondrial dysfunction and cell death.

ER stress

The endoplasmic reticulum (ER) is a major site for calcium storage, carbohydrate metabolism, protein synthesis and folding, and lipid and steroid synthesis [77]. Any imbalance in these processes or saturation of the ER membrane with lipid can lead to ER stress which activates the UPR to restore ER function. The three main UPR-mediated transmembrane proteins activated in ER stress are the serine/threonine-protein kinase/endoribonuclease inositol-requiring enzyme 1 (IRE1)/X-box binding protein 1 (XBP1), protein kinase R (PKR)-like endoplasmic reticulum kinase (PERK)/eukaryotic translation initiation factor 2α (eIF2α), and activating transcription factor-6 (ATF6). IRE1/X-box mediates cell apoptosis via the TNF receptor-associated factor 2 (TRAF2) and JNK activation while PERK and ATF6-mediated apoptosis includes activation of the pro-apoptotic C/emopamil binding protein (C/EBP) homologous protein (CHOP) [78]. In NASH, there is a decrease in ER efficiency but an increased demand for protein synthesis/folding/repair [10]. The result is ER stress in which there are misfolded or unfolded proteins inside the ER lumen. When this occurs, the UPR is initiated. Restoration of proper ER functioning requires multiple actions, including degradation of the unfolded or misfolded proteins, an increase in folding enzymes and chaperones, and translational inhibition [77]. However, in NASH there may be protracted activation of the UPR which can result in apoptosis and cell death [79, 80].

JNK, CHOP, BH3-only proteins and other cell death pathways

With lipotoxicity in hepatocytes, proteins that are involved in multiple pathways are upregulated including JNK, CHOP, PERK, BH3-only proteins, and caspases, ultimately resulting in mitochondrial dysfunction and apoptosis. When activated by either extracellular TNF-a or intracellular oxidative or ER stress, the MAP3 kinase apoptosis signal-regulating kinase 1 (ASK1) activates the p38/JNK pathway, leading to hepatocyte apoptosis and fibrosis [56]. Because ASK1 inhibition down regulates p38- and JNK-dependent signaling, it was proposed that inhibitors might have the potential to treat multiple inflammatory pathologies, including NASH [81]. Indeed, in a recent trial the ASK1 inhibitor selonsertib was shown to have the potential to reduce liver fibrosis in NASH patients [82]; it is currently being assessed in phase 3 studies. Wang et al recently reported that CASP8 and FADD-like apoptosis regulator (CFLAR) directly targets ASK1 by blocking its N-terminus-mediated dimerization, eventually blocking ASK1 activation of JNK1 [83]. In mice and monkey NASH models, they showed that a peptide fragment of the protein, CFLAR(S1), led to histological improvement including improvement in fibrosis. In another study, Zhang et al explored the potential mechanism underlying ASK1 activation [84]. They found that the deubiquitinase tumor necrosis factor alpha-induced protein 3 (TNFAIP3) plays a key role as an endogenous suppressor of ASK1 activation, and that TNFAIP3 directly interacts with and deubiquitinates ASK1 in hepatocytes. Hepatocyte-specific ablation in mice of Tnfaip3 led to exacerbation of NAFLD/NASH while TNFAIP3 gene delivery to the livers of mice and monkeys’ models of NASH blocked disease onset and progression. These new studies clearly show the important effect of ASK1 in NASH pathophysiology and the great potential for targeting ASK1 as a therapeutic strategy in NASH.

LPC also plays a main role in the induction of apoptotic pathways downstream of JNK or glycogen synthase kinase 3 (GSK3) activation and CHOP induction, ultimately resulting in the upregulation of pro-death proteins such as PUMA (a BH3-only protein) [85]. Other players in cell death pathways include the BCL-2 family proteins which play a key role in the permeabilization of the mitochondrial membrane. These proteins contain one or more Bcl-2 homology (BH) domains and have anti-apoptotic and pro-apoptotic members. The latter are divided among the multi BH domain members, including Bak, Bax and Bok, and the BH3-only proteins, including Bad, Bid, BIM, and PUMA [86]. Caspase-8 activation, mediated by Fas ligand binding to its receptor Fas, leads to the cleavage of Bid which subsequently activates Bak and Bax, all of which lead to formation of pores in the outer mitochondrial membrane. NASH patients were found to over-express the FAS receptor [87]. In addition, Bim and PUMA have been shown to play a role in FFA-lipotoxicity via caspase-3 activation and apoptosis, mitochondrial dysfunction and Bax activation that leads to mitochondrial permeability, causing activation of caspases 3, 6 and 7 resulting in DNA damage and cell death [88].

Hepatic Inflammation, Receptors, and the Immune System

Toll-like receptors (TLRs), nucleotide-binding oligomerization domain receptors (NOD-like) receptors (or NLRs) and recognition signal receptors are involved in innate immune system activation and play a role in NASH [89]. Their activation leads to a pro-inflammatory cytokine cascade, with recruitment of various immune cells such as macrophages and T cells which have been shown to induce insulin resistance and fatty liver development [90–92]. In NASH, intestinal permeability increases and the intestine-derived pathogen-associated molecular patterns (PAMPs) and LPS translocated to the liver lead to activation of TLRs [49, 93]. In particular, the activation of TLR2, TLR4, and TLR9 induces several cytokines including TGF-β, IL-1β and TNFα, which leads to lipid accumulation, HSC activation and hepatic apoptosis. This apoptosis leads to further activation of TLRs and further release of cytokines, including IL-6. TLRs 4 and 9 have been shown to induce hepatic inflammation and fibrosis through inflammasome activation. The effects of TLR4 are primarily driven by the NF-κB and mitogen-activated protein kinase (MAPK) cascades which regulate inflammation, insulin resistance and metabolic homeostasis [94–96]. TLR4 is internalized into an endosome network where it is sorted either to lysosomes for degradation or to the plasma membrane [97, 98]. TAK-242 and E5564 are TLR-4 inhibitors that have been shown to counter hepatic glucose production and insulin resistance [99]. A recent study found that membrane BAX inhibitor motif-containing 1 (TMBIM1) leads to lysosomal degradation of TLR4 and prevents further activation. TMBIM1 protected against NAFLD in mice and NASH progression in monkeys [100].

NLR activation results in assembly of the caspase 1-containing inflammasome, leading to inflammation and apoptosis. NOD1 and NOD2 have been associated with inflammatory diseases, and both NOD1 and NOD2 levels are highly expressed in the liver cells [90, 89]. For example, the NOD-like receptor protein 3 (NLRP3) inflammasome has been found to be activated in NAFLD. A new small molecule NLRP3 inhibitor, MCC950, has shown to reduce liver enzymes, inflammatory markers, histological inflammation and fibrosis in the MCD-diet model [101].

The Hedgehog pathway consists of a complex signaling cascade that is important for the immune response and is essential in embryogenesis but can be activated later in life in the setting of tissue regeneration [102]. In mice, Hedgehog pathway activation resulted from liver fat injury and was associated with steatohepatitis and fibrosis [103]. Similarly, in humans Hedgehog activity has been shown to be correlated with the severity of hepatocyte ballooning, portal inflammation and liver fibrosis [104]. Response to treatment with vitamin E in NASH was associated with decreased Hedgehog signaling [105]. It is thought that lipotoxicity can activate the Hedgehog pathway which in turn stimulates inflammatory responses, particularly natural killer T cells, and stimulates hepatocyte differentiation and HSC activation leading to fibrosis [103].

Nuclear Receptors

Nuclear receptors regulate glucose and lipid metabolism in the liver. They consist of seven families known as NR0-NR6. NR1 plays a particularly important role in NAFLD. It is located in the nucleus heterodimerized with the retinoid X receptor (RXR) and includes: NR1C1–3 (the peroxisome proliferator-activated receptors, PPARα), NR1H2–3 (the liver X receptors, LXR), NR1H4 (the farnesoid X receptor, FXR), NR1I2 (the constitutive androstane receptor, CAR), and NR113 (the pregnane X receptor, PXR). Here we focus on PPAR and FXR [106–108]. PPARα regulates ß-oxidation and cholesterol metabolism during the fasting state or when metabolism increases in adipose or muscle tissues [107]. PPARs inhibit inflammation in obesity by acting on NF-κB and AP1 transcription factors, regulating metabolism by inducing transcription of adiponectin (PPAR ) and fibroblast growth factor-21 (FGF21). Hepatic PPARα expression has been shown to be decreased in NAFLD leading to steatosis, but it increases after diet and exercise [109]. Animal models of NASH have shown improvement in steatosis if both PPAR ß/δ and PPARα are induced. PPAR activation also improves insulin resistance through various mechanisms [110]. Pioglitazone has been shown to improve histology in humans with NASH in the PIVEN trial and a recent study showed that elafibranor (an agonist of PPARδ and PPARα) led to improved histology in NASH[111].

FXR receptors are highly expressed in liver, intestine, kidneys and adrenals. They inhibit CYP7A1 and sterol 12-a-hydroxylase (CYP8B1), gene expressions which are involved in the metabolism of cholesterol to bile acids. FXR activation leads to the expression of various genes involved in glucose, lipid, and lipoprotein metabolism and bile acid synthesis [112]. Hepatic FXR inhibits fatty acid uptake and synthesis and stimulates B-oxidation. In animal models of NASH, FXR activation led to improvement in steatosis [113, 114]. Obeticholic acid, an FXR agonist, has been shown to improve histology with improvement in fibrosis in NASH patients [115]. Intestinal FXR may reduce weight, liver glucose production and steatosis, especially via stimulating human fibroblast growth factor-19 (FGF19) [116–118]. The administration of FGF19 in animal models increases free fat oxidation and decreases liver triglycerides and glucose levels. Further, intestinal FXR activation led to browning of the adipose tissue which in turn led to reduction in insulin resistance and obesity [119].

Tissue Regeneration and Fibrosis

Liver fibrosis is the accumulation of an excessive amount in the liver of extracellular matrix (ECM) proteins such as type I collagen. At later NASH stages, many types of cells play a role in this but the most important are liver myofibroblasts, originally from HSCs, portal fibroblasts (PFs) and mesothelial cells [120]. HSC activation is a result of the crosstalk between the previously mentioned pathways as well as inflammatory cells such as macrophages, NK cells, and lymphocytes [120]. Recent studies have shown correlation between the Notch signaling pathway and HSC activation [121–123]. TGF-β-activated kinase 1 (TAK1)/JNK, p38 pathways and high mobility group box 1 (HMGB1) released from necrotic hepatocytes lead to HSC activation. The ongoing tissue regeneration process counteracted by the above-mentioned pathways promotes first the shift from quiescent HSCs to an ECM-producing milieu in NASH and subsequently the undesired outcomes of fibrosis and cirrhosis.

Summary and Our Opinion

In our review, we have noted many recent findings on NASH pathogenesis that have expanded our understanding of its complexity. However, we think that it is very important to note that these pathophysiological pathways may vary among patients with NASH. For instance, de novo lipogenesis might play a major role in some patients while other pathways have a lesser role and vice versa. This opens the door for the possibility of the existence of many sub-types of NASH. Thus, effective therapy may require identifying a given patient’s subtype in order to choose which pathways should be targeted. For most patients, this is likely to mean targeting multiple pathways but the specific pathways that will need to be included in a treatment plan may vary between patients. Our observation has been supported by a recent study that showed differences in the pathophysiology in subtypes of NASH patients [124].

The findings of recent pathogenesis studies have provided key new insights that may ultimately lead to important therapeutic advances. We are already seeing clinically significant improvements in NASH patients treated with agents such as ACC, ASK1 and SCD1 inhibitors and FXR agonists. Further advancing our understanding of the mechanisms that underlie the pathogenesis of the disease and the complex interplay between them will be crucial for developing therapies that can effectively counter the multiple hits that lead to disease.