In the industrialized world, non-alcoholic fatty liver disease (NAFLD) is the most prevalent form of liver disease and sometimes precedes cirrhosis and hepatocellular carcinoma. A firm understanding of the mechanisms underlying the onset and progression of NAFLD are critical to combat the disease. The current research on NAFLD suggests a two-hit hypothesis. The first hit is caused by an accumulation of fat in hepatocytes which leads to steatosis. The second-hit is believed to be caused by inflammation and oxidative stress in the liver, leading to the progression from steatosis to nonalcoholic steatohepatitis and cirrhosis, only after the liver has been sensitized by the fat accumulation (Della Corte et al., 2012). This review will cover two mechanisms underlying the first hit of NAFLD.
As mentioned in the History and Metabolic Context section, the liver plays a critical role in the flux of free fatty acids and the transport of triglycerides. The first hit which is caused by an accumulation of fat in the liver results from a disruption in the balance between triglycerides into the liver and triglycerides transported out of the liver. Mutations in adipose triglyceride lipase (ATGL) which is one of the primary lipases in the liver, prevents the release of free fatty acids (FFAs) from lipid droplets leading to steatosis (Cohen et al., 2011). Mutations in hydroxyacyl-CoA transferases which are required for the oxidation of FFAs in the mitochondria also lead to increased fat accumulation in the liver (Cohen et al., 2011). Additionally, steatosis is caused by mutations in VLDL which transports triglycerides out of the liver and mutations in microsomal triglyceride transfer protein which adds a triglyceride to a lipoprotein particle (Cohen et al., 2011). While each of these genetic defects promote fat accumulation in hepatocytes, they are very rare variants and do not account for the one-third of Americans who are afflicted by NAFLD.
Identification of PNPLA3 Variants
A recent clinical study of over 2000 individuals titled the “Dallas Heart Study” found that hepatic steatosis was present in 45% of Hispanics, 33% of individuals of European ancestry, and 24% of African Americans (Cohen et al., 2011). The higher frequency of hepatic steatosis found in Hispanics was believed to be due to increased obesity and insulin resistance in that population, but researchers were unable to explain the decreased percentage of African Americans with hepatic steatosis using the same factors. With the same pool of individuals from the Dallas Heart Study, Romeo et al. performed a genome-wide association analysis (GWAS) to identify sequence variations in individuals with NAFLD (2008). The research team assayed single nucleotide polymorphisms in 3,383 African-Americans, Caucasians, and Hispanics. 2,111 of the participants underwent 1H-MRS, proton magnetic resonance spectroscopy, which non-invasively quantifies hepatic fat content (Romeo et al., 2008). From the GWAS study, Romeo et al. identified a single variant in PNPLA3 which was strongly associated with hepatic fat content and severity of NAFLD (2008). The SNP changes a cytosine to a guanine which results in a missense mutation that changes an isoleucine residue to methionine (Romeo et al., 2008). The highest frequency of the I148M mutation was in the Hispanic population which corresponds with the previous knowledge that hepatic steatosis is present at a higher frequency in Hispanics (Romeo et al., 2008). Additionally, another variant, S453I, was associated with lower hepatic fat content and commonly found in the African American population, but rarely found in European Americans and Hispanics (Romeo et al., 2008). The identification of the two PNPLA3 variants allowed for further study into the function of PNPLA3 and its role in NAFLD.
Structure and Activity of PNPLA3
PNPLA3, patatin-like phospholipase domain containing protein) belongs to the patatin and phospholipase superfamily. Patatin is a storage protein, but also contains lipid acyl hydrolase activity. The superfamily also consists of cytosolic phospholipase A2 which hydrolyzes the ester bond of phospholipids releasing arachadonic acid using a serine nucleophile (Dubuquoy et al., 2012). PNPLA3 has sequence similarity to PNPLA2 (otherwise known as adipose triglyceride lipase) which is a major triglyceride lipase in adipose tissue, but unlike PNPLA2, PNPLA3 is primarily expressed in the liver (Dubuquoy et al., 2012). Thus, the sequence similarity to PNPLA2 as well as the I148M variant correspondence to increased hepatic fat content suggests that PNPLA3 also functions as a triglyceride lipase, but in the liver (Dubuquoy et al., 2012).
Confirming this proposal, Hyysalo et al. found that in mice overexpressing human PNPLA3 I148M circulating long-chain triglycerides were depleted (2014). This is significant because triglycerides in lipid droplets are hydrolyzed before being reassembled and packaged into VLDLs for export from the liver. Since long-chain triglycerides are depleted in systemic circulation, it is likely that the PNPLA3 I148M impedes the hydrolysis of these triglycerides on the surface of lipid droplets and prevents their export from the liver, thus supporting the role of PNPLA3 in the hydrolysis of intrahepatocellular triglycerides. He et al. confirmed that PNPLA3 is associated with lipid droplets by ultracentrifugation and subsequent immunoblotting (2010).
Huang et al. purified human PNPLA3 from recombinant insect cells and found that the I148M substitution leads to an increase in Km and decrease in Vmax (2010). The research team also discovered that PNPLA3 has a strong preference for triglycerides containing oleic acid (18:1). He et al. developed a model of the patatin-like domain of PNPLA3 using the computer program SWISS-MODEL which constructed a homology model between human PNPLA3 and the x-ray structure of heartleaf horsenettle patatin (PDB:1OXW) (2010). The model of the PNPLA3 I148M mutant displayed that the Ile148 is adjacent to the residues Ser47 and Asp166 which form the catalytic dyad of the enzyme and the hydrophobic substrate-binding groove (He et al., 2010). When methionine is substituted for isoleucine, the longer methionine chain disrupts interaction between the substrate and the catalytic Ser47, leading to decreased hydrolysis of long-chain triglycerides. Together the structure and activity of PNPLA3 support the role of PNPLA3 in the hydrolysis of long-chain triglycerides and uphold the proposal that the I148M mutation leads to decreased enzymatic activity which causes increased hepatic fat accumulation.
Regulation of PNPLA3
PNPLA3 which functions as a triglyceride lipase in human hepatocytes is critical for maintaining liver fat homeostasis. Understanding the regulation of PNPLA3 is necessary to figure out how to target PNPLA3 in order to prevent hepatic fat accumulation. The observation that mouse PNPLA3 mRNA levels increase dramatically upon feeding and are elevated in obese, insulin-resistant mice suggested that PNPLA3 is regulated by insulin in some manner (Huang et al., 2010). The transcription factor SREBP-1c, as mentioned in History and Metabolic Context, is a major regulator of fatty acid biosynthetic pathways and is stimulated by insulin through activation of liver X receptors which bind to the promoter region of SREBP-1c. Upon stimulation, SREBP-1c is transported from the endoplasmic reticulum to the golgi where it is cleaved and released as an active transcription factor (Huang et al., 2010). In SREBP-1c knockout mice, PNPLA3 expression is decreased even when LXR agonists are added, signifying that PNPLA3 is directly regulated by LXR-activation of SREBP-1c (Huang et al., 2010). Additionally, when treated with LXR agonist, PNPLA3 mRNA levels significantly increased (Huang et al., 2010). Chromatin immunoprecipitation assays confirmed that SREBP-1c directly regulates PNPLA3 by identifying a response element in the first intron of PNPLA3 that SREBP-1c binds (Huang et al., 2010). These results confirm that SREBP-1c acts as a direct regulator of PNPLA3 expression through insulin-stimulated LXR activation.
Fructose’s Contribution to NAFLD
Although PNPLA3 variants have a significant effect on fact content among different ethnic populations, the Dallas Heart Study also concluded that hepatic steatosis is uncommon among lean individuals. From a study of over 2000 individuals, they concluded that 9% of lean individuals (BMI < 25kg/m2) had steatosis, compared to 51% of obese individuals (BMI > 35kg/m2) (Cohen et al., 2011). From these studies it is evident that obesity plays a significant role in NAFLD and since obesity is often caused by improper food consumption and lack of physical activity, researchers proposed that the increase in NAFLD over the past thirty years is likely due to changes in the quantity and composition of food (Cohen et al., 2011). Dietary fructose has been put in the limelight as one of the main culprits behind metabolic syndrome. Over the past century, dietary fructose consumption has increased from 15g/day to 75g/day (Lustig, 2013). According to Lustig, fructose is not needed for any biochemical reaction except in the production of semen, and fructose can even be produced from glucose via the aldose reductase/sorbitol pathway to make semen (2013). However, increased fructose does play a significant role in de novo lipogenesis in the liver and is a likely cause of fat accumulation in the liver.
One of the major problems with elevated dietary intake of fructose is that unlike glucose, fructose is taken up almost entirely by the liver (Cohen et al., 2011). Fructose enters hepatoctyes through glucose transporters, specifically GLUT2, GLUT5, and GLUT8. DeBosch et al. found that in GLUT8 knockout hepatocytes, fructose uptake was significantly decreased and de novo lipogenesis was also inhibited (2014). GLUT8 knockout mice fed a high fat diet had significantly lower Oil red-O staining density suggesting that GLUT8 is critical for developing fructose-induced steatosis (DeBosch et al., 2014). Additionally qRT-PCR analysis of GLUT8KO mice liver tissue revealed decreased fatty acid synthase and SREBP mRNA levels upon high fat diet feeding, displaying that fructose-induced de novo lipogenic genes were inhibited at the gene expression level by reducing fructose uptake into hepatoctyes (DeBosch et al., 2014). These results indicate that fructose uptake into hepatocytes is critical for the promotion of de novo lipogenesis and fructose-induced steatosis.
Understanding fructose metabolism is critical to unraveling how fructose is capable of inducing de novo lipogenesis. Upon entering hepatocytes, fructose is phosphorylated by the enzyme fructokinase to fructose-1-phosphate, unlike glucose which is primarily converted to glycogen under the influence of insulin signaling. Fructose-1-phosphate then undergoes glycolysis and produces pyruvate which is converted to acetyl-CoA. With significant consumption of fructose seen in the Western diet, excess acetyl-CoA is formed and acetyl-CoA is one of the primary substrates of fatty acid biosynthesis (Lustig 2013). Additionally, glyceraldehyde and dihydroxyacetonephosphate which are intermediaries of glycolysis can recombine to form fructose-1,6-bisphosphate which can then be converted to Xylulose-5-phosphate. Xylulose-5-phosphate powerfully stimulates protein phosphatase 2A which activates the transcription factor ChREBP (Lustig 2013). ChREBP then enhances the expression of ATP citrate lyase which converts citrate and coenzyme A to acetyl-CoA (Lustig 2013). ChREBP also increases the expression of acetyl-CoA carboxylase which forms malonyl-CoA, a substrate for fatty acid biosynthesis, and fatty acid synthase which is a multienzyme protein that catalyzes the production of palmitate from acetyl-CoA and malonyl-CoA (Lustig 2013).
Additionally, fructose can stimulate PPAR-γ, which activates SREBP-1c (Lustig 2013). The transcription factor SREBP-1c activates ATP citrate lyase, acetyl-CoA carboxylase, and fatty acid synthase, as well as enzymes responsible for the production of NADPH which is a critical cofactor of enzymes responsible for fatty acid biosynthesis (Horton et al., 2002). Therefore, fructose has been stated to drive “double de novo lipogenesis” because of its activation of both SREBP-1c and ChREBP (Lustig 2013). Additionally, studies revealed that when fructose and glucose enter hepatocytes simultaneously, glucose typically enters the glycogenic pathway which forces fructose to go down the lipogenic pathway (Lustig 2013). A recent human study displayed that fructose feeding increased de novo lipogenesis to 17% from 2% with glucose feeding (Lustig 2013). Therefore, excessive consumption of dietary fructose plays a critical role in the production of fatty acids in the liver and is an essential component to hepatic steatosis and the onset of NAFLD.
To learn about the treatment of NAFLD please go to the Treatments and Disease Management Section!