Mechanisms Underlying the Inflammatory Response in NASH
Our understanding of what causes NASH at the molecular level remains mostly speculative. The original 'two-hit' hypothesis as defined by Day and James proposed a prerequisite accumulation of lipid as the first event, while the 'second hit' was oxidative injury and subsequent necroinflammation. More recently, the theory of systemic lipotoxicity has been applied to NASH, where excessive or dysfunctional regulation of free fatty acids (FFAs) and/or their metabolites induces cellular injury and death.[1,27,42] This theory attempts to link dietary changes, overnutrition and dysfunctional lipid partitioning to insulin resistance, inflammation and the subsequent cascade of events that results in cellular injury. In this proposal, lipid accumulation in the hepatocyte is not a prerequisite but is a result of excess FFAs, and is thus protective through sequestering the more toxic FFAs or metabolites in the form of triglyceride.
There appears to be a reproducible connection between overnutrition, metabolic syndrome and NAFLD, and the most striking correlates are with visceral fat and insulin resistance. Normal weight yet metabolically obese individuals have also been described, but are also linked by overnutrition through reduced exercise and decreased exercise tolerance. A history of recent or progressive weight gain (even in the absence of clinical obesity) is very common in NASH, while increased exercise with modest weight loss is associated with normalizations in liver enzymes and histology. These observations unify the metabolic syndrome and NASH as diseases of overnutrition. A recent review by Tilg and Moschen proposed the 'multiple parallel hits' hypothesis, where inflammation arises as a consequence of many parallel hits originating from visceral adipose tissue and/or gut; according to this hypothesis, gut-derived bacterial byproducts, cytokine and adipokine signaling, endoplasmic reticulum (ER) stress and innate immunity emerge as key factors in NASH pathogenesis.
Fatty Acids & Fructose
Evidence for excess fatty acids playing a central role in the metabolic syndrome and NASH is growing. The primary sources of hepatic FFA are adipocyte triglyceride lipolysis or hepatocyte de novo lipogenesis, while dietary fat contributes approximately 15%. Circulating triglycerides are not a direct source of hepatic fat, and thus efforts to reduce serum triglyceride levels might not have a significant direct benefit on the liver.[48,49] As noted previously, polymorphisms in the PNPLA3 gene family are strongly correlated with NASH, thus supporting the vital role of visceral adipose lipolysis in hepatic fat delivery and the subsequent cascade resulting in NASH.[50,51] This theory is further supported by observations by Musso et al., who demonstrated that subjects with NASH had elevated serum FFA in response to a fatty meal, with an associated reduction in postprandial adiponectin.
Consumption of high-fat diets, especially saturated fat, has been linked to cardiovascular disease and diabetes, but their link to NASH is less well defined. Fat consumption has not been demonstrated to be able to consistently differentiate between simple steatosis and NASH patients.[53–55] Similarly, animal models of high-fat diet-induced insulin resistance and NAFLD are typically without hepatic inflammation.
Excess carbohydrate consumption or impaired peripheral glucose disposal results in de novo synthesis of fatty acids. Normally, de novo synthesis accounts for 5% of hepatic fat content; however, in subjects with NASH, up to 25% of the hepatic fat content may be due to de novo lipogenesis. Changes in dietary habits, including substantial increases in dietary fructose consumption over the past 30 years, may be partially responsible for the overwhelming production of saturated FFA. This is significant in that the metabolism of fructose is distinct from glucose, as phosphorylation by fructokinase is not rate limited, thus resulting in overwhelming production of saturated FFA. This theory is supported by the observations that animals fed a high-fructose diet develop hepatic steatosis and exacerbation of diet-induced NASH.[56–58] Recently, Ouyang et al. correlated dietary consumption and gene expression of fructokinase and fatty acid synthase in human subjects with NASH. Regardless of the cause, increased saturated FFAs are delivered to the liver where they are processed and may result in disease.
Lipids as Inflammatory Mediators
The derangement of lipid profile has been well described by Sanyal et al.[60,61] These studies have provided insights into whole-body lipid metabolism through a snapshot of the metabolic state of NASH. Contrary to once-conventional wisdom, they demonstrated that disturbances in lipid profiles may be much more complex than simple fluctuations in FFAs. Specifically, they demonstrated increased lipogenesis with an increased monounsaturated to saturated fatty acids ratio; increased diacylglycerol and free cholesterol; and depletion of arachidonic acid (the precursor of many biologically active eicosanoids). They speculated that increased lipogenesis, increased relative unsaturated fatty acids and triglyceride synthesis may be an adaptive response to fatty acid overload, which may allow for membrane stabilization and improved insulin sensitivity, again consistent with the theory of systemic lipotoxicity. The observed reduction of arachidonic acid probably reflects utilization. Selective increases in the byproducts of the lipoxygenase pathway but not cycloxygenase were also observed. NASH subjects had increased proinflammatory eicosanoids. Notably, 15-hydroxyeicosatetraenoic acid (a proinflammatory product of 15-lipoxygenase) was significantly increased in NASH; however, the levels of lipoxin A4 (an anti-inflammatory product of 15-lipoxygenase) were not increased. NASH was also associated with an increase in the levels of 11-hydroxyeicosatetraenoic acid, a nonenzymatic oxidation product of arachidonic acid. These observations may have major implications in the understanding of the proinflammatory state, and possibly the pathophysiology of NASH. Finally, evidence of impaired peroxisome metabolism was noted, the significance of which is unclear.
Visceral Adipose Tissue & Inflammatory Mediators
Visceral obesity is a hallmark of the metabolic syndrome and is almost invariably associated with NAFLD.[1,26] Visceral adipose tissue has also been demonstrated to secrete more proinflammatory cytokines, including TNF-α, IL-6 and monocyte chemoattractant protein-1, compared with subcutaneous adipose tissue, possibly through recruitment of innate immune cells.[62–67] Although highly variable between individuals, there is probably a 'critical threshold' for visceral fat; once the threshold is reached, inflammatory and hormonal signals promote insulin resistance together with its complications, including NASH. The recent discovery of the PNPLA3 family, including adiponutrin, may explain the link between visceral adipose tissue and NASH and the variable threshold between individuals.[18,20]
Adipose tissue is not solely a reservoir for excess nutrients, but is an active endocrine tissue. Visceral and subcutaneous adipose tissues secrete leptin and adiponectin among other hormones. The primary source of adiponectin is unclear, although growing evidence supports its protective effects through improving hepatic insulin resistance and mobilization of visceral and hepatic fat.[68,69] Low serum adiponectin has been identified in NASH subjects, and its overexpression is protective in animal models of NAFLD.[49,51,70] Increased adiponectin production in response to thiazolidinedione treatment appears to occur primarily through visceral adipose tissue. Thiazolidinedione treatment has been demonstrated to reduce visceral and hepatic fat deposition with an associated increase in subcutaneous fat, which is probably mediated through increased mobilization of visceral and hepatic fat content.[72–75] Similarly, thiazolidinedione treatment-associated improvement in histological NASH was strongly correlated with increased adiponectin. Likewise, overproduction of adiponectin resulted in shunting visceral fat to subcutaneous adipose tissue in a mouse model of NAFLD. Finally, Musso et al. demonstrated a correlation of reduced postprandial adiponectin with elevated serum FFAs in response to a fatty meal. Follow-up studies have demonstrated a strong correlation between low serum adiponectin concentrations and NASH.[50,51]
Role of Gut-derived Bacteria & Flora
The role of gut-derived bacteria and their products in NAFLD is unclear. Several studies have demonstrated a correlation between small intestinal bacterial overgrowth and NAFLD;[77,78] however, there is no clear differentiation between NAFLD and NASH. As noted previously, NASH is associated with an increase in TNF-α, but an increase in endotoxin or bacterial byproducts has not been observed. Supporting the theory, Li et al. have demonstrated a protective effect of both probiotics and anti-TNF-α antibodies in a mouse model of NAFLD, resulting in improved hepatic insulin resistance and hepatic steatosis.
Oxidative stress occurs in NASH, but its exact role in disease pathophysiology has not been clearly established.[80,81] The liver is well-endowed with the antioxidant glutathione and metabolic pathways for dealing with reactive oxygen or nitrogen species that occur as a result of oxidation of lipids. Depletion of hepatic antioxidants has not been reliably demonstrated in NASH, suggesting that this pathway is relatively well-adapted to handle the byproducts of fatty acid metabolism.[82,83]
Oxidative stress may directly activate an immune response and, subsequently, drive further inflammation, or may be the result of inflammation. Hepatic oxidative stress, lipid peroxidation and ER stress can directly activate the inhibitor of NF-κB kinase or JNK to activate transcription of proinflammatory cytokines. Likewise, oxidative stress-dependent antibodies have been associated with advanced fibrosis or cirrhosis, suggesting that an oxidative stress-induced immune response is associated with disease progression. Further evidence supporting the role of oxidative stress in the pathogenesis of NASH was provided by the Pioglitazone vs Vitamin E vs Placebo for Treatment of Non-Diabetic Patients With Nonalcoholic Steatohepatitis (PIVENS) trial, where vitamin E was associated with improvement in histological NASH.
Expert Rev Gastroenterol Hepatol. 2011;5(2):189-200. © 2011 Expert Reviews Ltd.