Elovl6 promotes nonalcoholic steatohepatitis


  • Potential conflict of interest: Nothing to report.

  • Supported by Grant-in-Aid for Scientific Research 21390275 (to H. S.), 22117502 (to H. S.), and 21689025 (to T. M.) from the Ministry of Science, Education, Culture, and Technology of Japan; the Kowa Life Science Foundation (T. M.); and Suzuken Memorial Foundation International (T. M.).


Nonalcoholic steatohepatitis (NASH) is associated with obesity and type 2 diabetes, and an increased risk for liver cirrhosis and cancer. ELOVL family member 6, elongation of very long chain fatty acids (Elovl6), is a microsomal enzyme that regulates the elongation of C12-16 saturated and monounsaturated fatty acids (FAs). We have shown previously that Elovl6 is a major target for sterol regulatory element binding proteins in the liver and that it plays a critical role in the development of obesity-induced insulin resistance by modifying FA composition. To further investigate the role of Elovl6 in the development of NASH and its underlying mechanism, we used three independent mouse models with loss or gain of function of Elovl6, and human liver samples isolated from patients with NASH. Our results demonstrate that (1) Elovl6 is a critical modulator for atherogenic high-fat diet–induced inflammation, oxidative stress, and fibrosis in the liver; (2) Elovl6 expression is positively correlated with severity of hepatosteatosis and liver injury in NASH patients; and (3) deletion of Elovl6 reduces palmitate-induced activation of the NLR family pyrin domain-containing 3 inflammasome; this could be at least one of the underlying mechanisms by which Elovl6 modulates the progress of NASH. Conclusion: Hepatic long-chain fatty acid composition is a novel determinant in NASH development, and Elovl6 could be a potential therapeutic target for the prevention and treatment of NASH. (HEPATOLOGY 2012;56:2199–2208)

Nonalcoholic fatty liver disease (NAFLD), characterized by triglyceride (TG) accumulation in hepatocytes (hepatosteatosis) in the absence of chronic alcohol consumption, is the most common cause of abnormal liver function.1 NAFLD is often associated with obesity, dyslipidemia, and insulin resistance (IR).2-4 A large majority of obese and diabetic subjects develop hepatosteatosis, and its prevalence has increased markedly among adults and children.3 Furthermore, nonalcoholic steatohepatitis (NASH), a severe form of NAFLD accompanied by cellular damage, inflammation, and fibrosis, can progress to cirrhosis and hepatocellular carcinoma.5, 6

The pathophysiological mechanisms underlying steatosis–NASH progression remain poorly understood, but a “two-hit” hypothesis has been proposed.7 The first hit is the initial hepatic lipid accumulation, but a second hit is required for liver injury, inflammation, and fibrosis. One important aspect of this hypothesis is that steatosis per se is not causal in NASH development; rather, it sensitizes the liver to the damaging effects of second hits so that stressors innocuous to a healthy liver lead to NASH development in a steatotic liver.8 Lipidomic analysis of NAFLD livers suggests that hepatic fat deposition and, more specifically, the type of fat deposited directly damage the liver and precipitate NASH development.9

Fatty acid (FA) composition of the lipid species could be another determinant of the development of hepatic IR and energy metabolism accompanying intracellular lipid accumulation. ELOVL family member 6 (Elovl6) is a microsomal enzyme involved in the elongation of saturated and monounsaturated FAs with 12, 14, and 16 carbons.10, 11 Loss of Elovl6 function reduces stearate (C18:0) and oleate (C18:1n-9) levels, and increases palmitate (C16:0) and palmitoleate (C16:1n-7) levels.12 Mice with targeted disruption of Elovl6 (Elovl6−/−) are protected against hepatic IR development when fed a high-fat/high-sucrose diet despite hepatosteatosis and obesity in these mice and wild-type mice being similar, suggesting that hepatic FA composition, particularly C16:0 to C18:0 conversion, is crucial for insulin sensitivity rather than mere lipid accumulation.12 Inhibiting this elongase might be a potential therapeutic target for NASH; this concept was tested in this study.

Materials and Methods

Animals and Diets.

Elovl6−/− mice were generated as described.12 Transgenic mice expressing human ELOVL6 under the control of the human α1 antitrypsin promoter were established. Briefly, an expression plasmid-based vector containing α1 antitrypsin promoter and human ELOVL6 complementary DNA, followed by a 3′ polyadenylation signal from the rabbit β-globin gene, was constructed. The resulting construct was microinjected into pronuclear oocytes of C57BL/6J mice. Germline transmission was confirmed via Southern blot analysis. An atherogenic high-fat (AHF) diet was prepared with a modification of a high-fat, high-sucrose diet with 1.25% cholesterol and 0.5% cholate (Oriental Yeast, Tokyo, Japan)13 (Supporting Table 1). Eight-week-old male C57BL/6J mice were purchased from CLEA Japan. Mice were housed in colony cages with a 12-hour light/dark cycle and given free access to food and water. We sacrificed mice during the early light phase in a nonfasted state. All animal husbandry and experiments were performed in compliance with the guiding principle of the University of Tsukuba and were approved by the Animal Experiment Committee of the University of Tsukuba.

Preparation of Recombinant Adenovirus and Animal Transduction.

Adenoviruses expressing full-length mouse Elovl6 (Ad-Elovl6) and GFP (Ad-GFP) control were prepared as described.12 Adenoviruses were injected into the mice via the tail vein at 1 × 109 plaque-forming units. Mice were sacrificed for analysis on day 7 after viral injection.

Human Samples.

Human studies were conducted in accordance with the Declaration of Helsinki and were approved by the institutional review boards of the University of Tsukuba and Kyusyu University Hospital. All subjects provided informed consent prior to their participation. Liver samples were obtained via biopsy from 14 patients with histologically diagnosed NASH who were admitted to Kyushu University Hospital. Normal liver was obtained via biopsy from 64 controls, whose liver function tests and histological findings were completely normal. Real-time quantitative PCR was performed using an ABI Prism 7300 instrument and primer probe sets from ABI. The expression levels of RNA were normalized to those of β-2 microglobulin RNA.

Isolation and Culture of Mouse Primary Hepatocytes.

Primary hepatocytes were isolated from male wild-type (WT) or Elovl6−/− mice by enzyme-based tissue digestion method as described.12 Cells were resuspended in William's Medium E (Gibco) containing penicillin and streptomysin supplemented with 10% fetal bovine serum and 30 nM dexamethasone before being plated onto collagen-coated dishes. After incubation for 4 hours to allow attachment, the medium was replaced with serum-free William's Medium E containing bovine serum albumin alone or palmitic acid coupled with bovine serum albumin (500 or 750 μM) with or without adenovirus.

Statistical Analysis.

Results are expressed as the mean ± SEM of at least three independent experiments. Data between groups were analyzed via Student t test or two-way analysis of variance followed by Tukey's procedure. P < 0.05 was considered significant.


4-HNE, 4-hydroxy-2-nonenal; 8-OHdG, 8-hydroxy-2′-deoxyguanosine; αSMA, α smooth muscle actin; Ad, adenovirus; AHF, atherogenic high-fat; ALT, alanine aminotransferase; Elovl6, ELOVL family member 6; FA, fatty acid; H&E, hematoxylin and eosin; IL-1β, interleukin 1β; IR, insulin resistance; JNK, c-Jun N-terminal kinase; LD, lipid droplet; mRNA, messenger RNA; NAFLD, nonalcoholic fatty liver disease; NASH, nonalcoholic steatohepatitis; NF-κB, nuclear factor κB; NLRP3, NLR family pyrin domain-containing 3; PPAR, peroxisome proliferator-activated receptor; qPCR, quantitative real-time polymerase chain reaction; ROS, reactive oxygen species; T-Cho, total cholesterol; Tg, transgenic; TG, triglyceride; WT, wild-type.


Elovl6 Deletion Protects Against AHF Diet-Induced Liver Injury Without Decreasing Hepatic Lipid Content.

Wild-type and Elovl6−/− mice were fed a chow or an AHF diet, a useful dietary model for progressive NASH.13 A 12-week AHF diet increased weight significantly in whole body (Fig. 1A), liver (Fig. 1B), and white adipose tissue (data not shown) in both WT and Elovl6−/− mice. Hepatic TG and total cholesterol (T-Cho) levels were markedly increased in both WT and Elovl6−/− mice (Fig. 1C). Serum levels of alanine aminotransferase (ALT), a specific marker for hepatic parenchymal injury, were markedly increased in AHF WT mice (Fig. 1D). These data demonstrated that AHF diet feeding reflected the etiology of NASH. Conversely, the AHF Elovl6−/− mice exhibited significantly lower serum ALT levels than the AHF WT mice. FA composition identified whether lack of Elovl6 modifies the hepatic fatty acyl profile (Fig. 1E). Consistent with a hepatic role of Elovl6, we found increased C16:0 and decreased C18:0 in the chow-fed Elovl6−/− mice versus WT mice.13 The AHF diet caused no significant difference in C16:0, C18:0, or C18:1n-9 FA composition between WT and Elovl6−/− mice.

Figure 1.

Elovl6−/− mice are protected against AHF diet–induced liver injury without ameliorating hepatosteatosis. Body weight (A), liver weight (B), hepatic TG and T-Cho levels (C), plasma ALT levels (D), and hepatic FA composition (E) of WT and Elovl6−/− mice fed a normal chow or AHF diet for 12 weeks (n = 14-15 per group). *P < 0.05, **P < 0.01.

The AHF WT mice exhibited a robust elevation in plasma insulin and a slight increase in plasma glucose levels, indicating their IR emergence (Supporting Fig. 1A,B). However, the AHF Elovl6−/− mice showed a significant reduction in plasma insulin compared with the AHF WT mice. Intraperitoneal glucose tolerance tests revealed that the Elovl6−/− mice had significantly better glucose tolerance than the WT mice (Supporting Fig. 1C). The AHF Elovl6−/− mice were more insulin-sensitive, as indicated by faster glucose clearance in the insulin tolerance test (Supporting Fig. 1D). These results are consistent with our previous report.12

Elovl6-Deficient Mice Are Protected from AHF Diet–Induced Inflammation, Oxidative Stress, and Fibrosis.

To examine a morphological change in the liver, we performed histological analysis. Wild-type mice on a 24-week AHF diet developed massive lipid deposition, hepatosteatosis, hepatocyte ballooning, and lobular infiltration of inflammatory cells, including lymphocytes and neutrophils (Fig. 2A). Immunohistochemical staining for 4-hydroxy-2-nonenal (4-HNE) and sirius red also revealed exacerbated oxidative damage and substantial collagen deposition in AHF WT mice. Interestingly, AHF Elovl6−/− mice showed slightly smaller lipid droplets (LDs) and attenuated inflammatory cell infiltration compared with AHF WT mice, whereas hepatic TG and T-Cho levels were similar in both groups (Fig. 1C). AHF Elovl6−/− mice had a significantly lower hepatic lobular inflammatory grade (evaluated by the number of inflammatory foci in hematoxylin and eosin [H&E]-stained liver sections) than AHF WT mice (Fig. 2B). Elovl6−/− mice were also more resistant to AHF diet–induced hepatic oxidative stress and exhibited significantly reduced hepatic accumulation of 4-HNE and 8-hydroxy-2′-deoxyguanosine (8-OHdG) than AHF WT mice (Fig. 2C). Moreover, consistent with histology stained with sirius red, significantly reduced hydroxyproline levels of the AHF diet–fed Elovl6−/− versus WT mice support their reduced hepatic fibrosis (Fig. 2D). These results suggest that Elovl6 deficiency abrogated the development of diet-induced hepatic inflammation, oxidative damage, and fibrosis, despite comparable hepatosteatosis in the Elovl6−/− and WT mice.

Figure 2.

Attenuated hepatic inflammation, oxidative stress, and fibrosis in AHF Elovl6−/− mice. (A) Representative H&E staining, immunostaining for 4-HNE, and sirius red staining in liver sections from WT and Elovl6−/− mice fed a normal chow or the AHF diet for 24 weeks. (B) The number of inflammatory foci per 20× original magnification were counted on H&E-stained sections from each animal (n = 11-14 per group). (C, D) Liver 8-OHdG (C) and hydroxyproline (D) levels in WT and Elovl6−/− mice fed a standard chow or the AHF diet for 24 weeks (n = 10-18 per group). *P < 0.05, **P < 0.01.

To understand the mechanisms by which Elovl6 is involved in NASH progression, we performed quantitative real-time polymerase chain reaction (qPCR). During the middle stage of steatohepatitis development, expression of fatty acid synthesis genes, including sterol regulatory element binding protein 1c (Srebf1c) and stearoyl-coenzyme A desaturase-1 (Scd1), increased and that of Elovl6 and fatty acid synthase (Fasn) decreased slightly in the livers of the AHF WT mice (Fig. 3A). Expression of Srebf1c and Fasn was not different, but Scd1 was repressed in the livers of AHF Elovl6−/− mice versus AHF WT mice (Fig. 3A). The expression levels of mitochondrial fatty acid β-oxidation genes such as peroxisome proliferator-activated receptor α (Ppara) and carnitine palmitoyltransferase 1a (Cpt1a) were similar between the AHF diet–fed WT and Elovl6−/− mice (Fig. 3A). Expression of genes for cholesterol synthesis, uptake, and efflux, such as sterol regulatory element-binding protein-2 (Srebf2), 3-hydroxy-3-methylglutaryl-coenzyme A synthase 1 (Hmgcs1), LDL receptor (Ldlr), and ABCA1 (Abca1), were coordinately repressed and elevated, respectively, in both groups of mice (Fig. 3A). In WT mice, expression of peroxisome proliferator-activated receptor γ (Pparg) and cell death–inducing DFFA-like effector c (Cidec), which promote TG accumulation in LDs, were induced significantly by the AHF diet (Fig. 3B). Strikingly, the Elovl6−/− mice displayed markedly lower expression and a complete absence of Pparg and Cidec induction by the AHF diet. Inflammatory response genes such as tumor necrosis factor a (Tnfa), chemokine (C-C motif) ligand 2 (Ccl2), and toll-like receptor 4 (Tlr4) were dramatically upregulated in the livers of AHF WT mice (Fig. 3B). However, expression of these genes decreased significantly in AHF Elovl6−/− mice. The expression levels of genes for the NADPH oxidase complex, neutrophil cytosolic factor 1 (Ncf1) and neutrophil cytosolic factor 2 (Ncf2), an important source of reactive oxygen species (ROS), were coordinately elevated in AHF WT mice, but decreased significantly in AHF Elovl6−/− mice (Fig. 3B). The expression levels of genes for transforming growth factor beta 1 (Tgfb1), collagen type I α1 (Col1a1), and collagen type IV α1 (Col4a1), critical inducers of fibrogenesis, were dramatically upregulated in the AHF WT mice yet suppressed in the AHF Elovl6−/− mice (Fig. 3B).

Figure 3.

Effects of Elovl6 deficiency on hepatic mRNA and protein levels involved in steatohepatitis. Wild-type and Elovl6−/− mice were fed a standard chow or AHF diet for 12 weeks (n = 12-15 per group). Livers were collected for gene expression and immunoblot analyses. (A, B) qPCR of genes for FA synthesis, β-oxidation, cholesterol metabolism (A), LD formation, inflammation, ROS generation, and fibrogenesis (B). (C) Immunoblotting for phosphorylated and total JNK, phosphorylated and total NF-κB, α-SMA, and α-tubulin. (D) Ratio between phosphorylated and total JNK, phosphorylated and total NF-κB, α-SMA and α-tubulin via densitometry analysis (n = 6 per group). *P < 0.05, **P < 0.01.

Decreased hepatic inflammation, oxidative stress, and fibrosis in the Elovl6−/− mice was further examined by measuring stresses and proinflammatory pathways. The AHF Elovl6−/− mice showed significantly decreased c-Jun N-terminal kinase (JNK) and nuclear factor κB (NF-κB) phosphorylation than the AHF diet–fed WT mice (Fig. 3C,D). α-Smooth muscle actin (α-SMA) levels decreased significantly in AHF Elovl6−/− mice (Fig. 3C, D).

Overexpression of Elovl6 in Liver Induces Inflammation, Oxidative Stress, Fibrogenesis, and Liver Injury.

To investigate whether hepatic Elovl6 overexpression can conversely induce steatohepatitis, we administered an adenovirus carrying full-length Elovl6 (Ad-Elovl6) or green fluorescent protein (Ad-GFP) into C57BL/6J mice. Elovl6 messenger RNA (mRNA) and activity increased up to 16-fold and 5.3-fold in the livers of Ad-Elovl6 mice versus Ad-GFP mice (Supporting Fig. 2A and Fig. 4A). Ad-Elovl6 mice exhibited significantly increased liver weight (Fig. 4B), hepatic TG (Fig. 4C), and serum ALT levels (Fig. 4D). Elovl6 overexpression changed the hepatic FA profile by significantly decreasing the amount of C16:0, C16:1n-7, C18:2n-6, and C20:5n-3, and increasing the amount of C18:0, C18:1n-9, and C20:3n-6 (Fig. 4E and Supporting Tables 7-10). Histological examination revealed a slight increase in liver TG and prominent liver damage in Ad-Elovl6 mice compared with Ad-GFP mice (Fig. 4F). Marked vascular congestion and neutrophil infiltration were observed in Ad-Elovl6 mice. Inflammation in Ad-Elovl6 mice was further confirmed by a higher inflammatory score than that of Ad-GFP mice (Supporting Fig. 2B). Ad-Elovl6 mice also showed increased hepatic oxidative stress, as indicated by a significant increase in hepatic 8-OHdG accumulation versus Ad-GFP mice (Supporting Fig. 2C).

Figure 4.

Elovl6 overexpression in livers of mice increases hepatic inflammation and liver injury. Nine-week-old male C57BL/6J mice were injected with Ad-GFP (n = 14) or Ad-Elovl6 (n = 15). On day 7 after virus injection, mice were sacrificed in a nonfasted state and livers were collected for analysis. (A-E) Hepatic palmitory-CoA elongation activity (A), liver weight (B), hepatic TG (C), plasma ALT levels (D), and hepatic FA composition (E) in Ad-GFP and Ad-Elovl6 mice (n = 4 per group). *P < 0.05, **P < 0.01. (F) Representative H&E-stained sections of livers from Ad-GFP and Ad-Elovl6 mice.

mRNA abundance for the lipogenic genes Srebf1c, Fasn, and Scd1 decreased in Ad-Elovl6 mice (Fig. 5A). Because lipogenic gene expression was similar or significantly increased in the livers of Ad-Elovl6 mice in comparison with the livers of Ad-Elovl6 mice on day 3 and day 5 after virus injection (Supporting Fig. 2D), this ameliorating effect on lipogenic gene expression may be attributable to severe liver injury in the Ad-Elovl6 mice. Ppara expression decreased in Ad-Elovl6 mice, but Cpt1a expression was similar between Ad-Elovl6 and Ad-GFP mice (Fig. 5A). Expression of genes for LD formation (Cidec), inflammation (Tnfa, Ccl2, Tlr4), oxidative stress (Ncf1 and Ncf2), and fibrosis (Tgfb1, Col1a1, Col4a1) were significantly increased in the livers of Ad-Elovl6 mice (Fig. 5B,C). Ad-Elovl6 mice showed significantly increased JNK phosphorylation and α-SMA levels than Ad-GFP mice (Fig. 5D,E). Thus, Elovl6 overexpression was sufficient to induce hepatic lipid accumulation, inflammation, oxidative stress, and fibrogenesis.

Figure 5.

Effects of Elovl6 overexpression on mRNA and protein levels involved in steatohepatitis. (A-C) qPCR analysis of genes for FA synthesis and β-oxidation (A), LD formation (B), inflammation, ROS generation, and fibrogenesis (C) in the livers of Ad-GFP and Ad-Elovl6 mice (n = 14-15 per group). (D) Immunoblotting for phosphorylated and total JNK, α-SMA, and α-tubulin in livers of Ad-GFP and Ad-Elovl6 mice. (E) Ratio between phosphorylated and total JNK, α-SMA and α-tubulin via densitometry analysis (n = 6 per group) (E). P < 0.05, **P < 0.01.

In FA composition analysis, Elovl6 overexpression induced the accumulation of hepatic C18:0 and C18:1n-9 (Fig. 4E). To distinguish an effect of C18:0 and C18:1n-9 in Ad-Elovl6 mice, a Scd1-specific RNA interference (RNAi; Scd1i) adenovirus was coinfected with Ad-Elovl6 to reduce the amount of C18:1n-9 and increase that of C18:0 (Supporting Fig. 3). The Ad-Elovl6 and Ad-Elovl6+Scd1i mice showed similar Elovl6 mRNA levels (Supporting Fig. 3A). Ad-Scd1i reduced hepatic Scd1 expression by 90% (Supporting Fig. 3A). The amount of C18:0 significantly increased and that of C18:1n-9 significantly decreased in Ad-Elovl6+Scd1i mice than in the Ad-Elovl6 mice (Supporting Fig. 3B). However, liver weight, hepatic TG, T-Cho, and serum ALT levels increased in association with Ad-Elovl6 overexpression were not affected by Scd1 knockdown (Supporting Fig. 3C-E). Hepatic expression of Tnfa, Ncf1, Ncf2, Tgfb1, Col1a1, which was induced by Elovl6 activation, remained unchanged by Scd1 knockdown (Supporting Fig. 3F). Inversely, Elovl6 activation-induced liver damage was not affected by coinfection of the Scd1-expressing adenovirus with Ad-Elovl6 (Supporting Fig. 4). These data indicate that Elovl6-regulated FA chain elongation from C16 to C18 is important and that SCD1-mediated C18:0 desaturation does not contribute to development of hepatic inflammation, oxidative stress, or fibrosis.

Transgenic Activation of Elovl6 Facilitates AHF Diet–Induced Steatohepatitis and Liver Injury.

To further examine chronic effects of activated Elovl6 on the liver, we created transgenic (Tg) mice overexpressing human ELOVL6 in the liver under the control of the human α1 antitrypsin promoter (Supporting Fig. 5A). Transgene expression and function were confirmed via northern blotting (Supporting Fig. 5B), qPCR (data not shown), and elevated elongation activity in liver homogenates (Supporting Fig. 5C). Consistent with enhanced Elovl6 activity, hepatic FA composition of C16:0 and C16:1n-7 was reduced, whereas that of C18:0 was increased in the chow-fed ELOVL6-Tg mice versus WT mice (Fig. 6A and Supporting Table 11). Under an AHF diet, levels of C16:1n-7 were significantly reduced in ELOVL6-Tg mice, but those of C16:0, C18:0, and C18:n-9 were indistinguishable between WT and ELOVL6-Tg mice. Standard chow diet–fed ELOVL6-Tg mice had no obvious abnormalities (Supporting Fig. 5D-L). However, ELOVL6-Tg mice fed the AHF diet showed markedly decreased body weight (Supporting Fig. 5D), much higher serum ALT levels, and hepatic lipid accumulation (Fig. 6B,C) than the WT mice. Massive LDs and prominent neutrophil infiltration were observed on H&E-stained liver samples of AHF ELOVL6-Tg mice (Fig. 6D). Western blotting revealed that the AHF diet induced higher JNK phosphorylation and more prominent α-SMA induction in the livers of ELOVL6-Tg mice versus WT mice (Fig. 6E). Taken together, hepatic overexpression of Elovl6 facilitated FA chain elongation from C16 to C18 and enhanced AHF diet–induced steatohepatitis.

Figure 6.

Genetic overexpression of ELOVL6 in hepatocytes accelerates AHF diet–induced steatohepatitis and liver injury. Fourteen- to 16-week-old WT and ELOVL6 transgenic (Tg) mice were fed chow or an AHF diet for 2 weeks (n = 8-11 per group). (A-C) Hepatic FA composition (A), serum ALT levels (B), and hepatic TG and T-Cho contents (C) of WT and ELOVL6-Tg mice fed a chow or AHF diet. (D) Representative H&E-stained liver sections of WT and ELOVL6-Tg mice fed an AHF diet. (E) Immunoblotting for phosphorylated and total JNK, α-SMA, and α-tubulin in livers of WT and ELOVL6-Tg mice fed a chow or AHF diet (n = 6 per group). *P < 0.05, **P < 0.01.

Increased Hepatic Elovl6 Expression in Human NASH.

To examine the possible association of hepatic ELOVL6 levels in human pathology, we analyzed liver biopsy samples of NASH patients. The clinical data are presented in Supporting Table 11. Human NASH liver exhibited an approximately 2-fold increase in ELOVL6 mRNA levels (Fig. 7A) accompanied by significant increases in SCD and SREBF1 mRNA and a dramatic decrease in IRS2 mRNA levels (Supporting Fig. 6A) compared with control subjects. Importantly, a significant correlation was observed between hepatic ELOVL6 expression and the percentage of steatosis, further linking NASH statues to high ELOVL6 expression in the human liver (Fig. 7B). The expression levels of SCD, SREBF1, and IRS2 showed no correlation with the percentage of steatosis (Supporting Fig. 6B-D). We then analyzed a correlation between ELOVL6 mRNA levels and physiological parameters and found a positive correlation between hepatic ELOVL6 expression and serum ALT levels (Fig. 7C). These observations suggest a potential role for hepatic ELOVL6 in the pathogenesis of human NASH.

Figure 7.

Hepatic expression of ELOVL6 is increased in NASH patients and correlates with steatosis and ALT levels. (A) qPCR analysis of ELOVL6 in the livers of control (n = 64) and NASH (n = 15) patients. (B) Correlation of the steatosis level with the hepatic expression of ELOVL6 in NASH patients. (C) Correlation of hepatic ELOVL6 expression with serum ALT (C) levels in control and NASH patients. **P < 0.01.

Elovl6 Plays an Important Role in Palmitate-Induced Inflammasome Activation in Hepatocytes.

The NLR family pyrin domain-containing 3 (NLRP3) inflammasom has emerged as a major contributor to inflammation in a variety of metabolic diseases including NASH.14, 15 Moreover, recent studies have shown that the saturated FA palmitate activates the inflammasome and induces interleukin 1β (IL-1β) release in macrophage and hepatocyte.16, 17 Because palmitate is a substrate of Elovl6, we tested whether palmitate signaling through Elovl6 leads to the activation of NLRP3 inflammasome. We found that the expression of genes for the NLRP3 inflammasome, including Nlrp3, PYD and CARD domain containing (Pycard), caspase-1 (Casp1), and interleukin-1β (Il1b), was significantly increased in the livers of AHF WT mice versus chow-fed WT mice (Fig. 8A). Dietary induction of these genes was significantly suppressed in Elovl6−/− mice. Conversely, hepatic overexpression of Elovl6 significantly increased expression of inflammasome genes (Fig. 8B).

Figure 8.

Elovl6 modulates NLRP3 inflammasome in the liver and in hepatocytes. (A, B) qPCR of genes for NLRP3 inflammasome in the livers of WT and Elovl6−/− mice fed a chow or AHF diet (A) and in Ad-GFP and Ad-Elovl6 mice (B). (C, D) NLRP3 inflammasome mRNA expression (C) and IL-1β levels released in supernatants (D) of primary hepatocytes isolated from WT and Elovl6−/− mice treated with or without palmitate (PA) (750 μM) for 24 hours (n = 4-6). (E, F) NLRP3 inflammasome mRNA expression (E) and IL-1β levels released in supernatants (F) of hepatocytes isolated from C57BL/6 mice infected with Ad-GFP or Ad-Elovl6 treated with or without PA (500 μM) for 24 hours (n = 4-6). *P < 0.05, **P < 0.01.

We next evaluated whether Elovl6 plays a role during inflammasome activation by addition of palmitate in hepatocytes. Expression of genes for inflammasome (Nlrp3, Pycard, Casp1), inflammation (Il1b, Tnfa), and IL-1β secretion were significantly increased by palmitate loading, but these effects were significantly suppressed in Elovl6−/− hepatocytes (Fig. 8C, D). In contrast, palmitate-induced inflammasome activation was accelerated by Elovl6 overexpression (Fig. 8E,F). Taken together, these results suggest diet-induced NLRP3 inflammasome activation can be modified by the Elovl6 activity. FA metabolites generated through Elovl6-mediated chain elongation in hepatocytes could modulate NLRP3 inflammasome activation, triggering inflammation in NASH.


Using three different animal models (gene disruption and adenoviral and transgenic overexpression), we demonstrate clearly that Elovl6 plays an important role in promoting NASH development. Elovl6 expression is critical for the main NASH-related pathologies: hepatic oxidative stress, inflammatory damage, liver injury, and fibrosis. According to the two-hit hypothesis, the initial hit is hepatic lipid accumulation, whereas the second hit is oxidative stress and inflammation.7 Whereas FA synthesis normally is not very active and may not be a major contributor to steatosis and NASH, this study seems to imply that fatty acid synthesis (elongation of C16 to C18 FAs) determines the first hit, the subsequent cellular stresses, and whether hepatocytes are damaged. Experiments with the combination of Elovl6 and Scd1i adenoviruses indicated that Elovl6-regulated FA chain elongation from C16 to C18 plays more important role rather than the subsequent Scd1-mediated C18 desaturation.

The Elovl6−/− mice are unique in that NASH pathologies were attenuated without amelioration of hepatosteatosis. Although large epidemiological studies suggest TG accumulation might negatively affect hepatosteatosis,18 recent evidence has indicated that TG might exert protective functions for removing potentially lipotoxic FA derivatives by channeling FA substrates into a storage form of TG.19-21 Thus, it is not contradictory even if hepatosteatosis is dissociated from NASH pathologies in Elovl6−/− mice. However, Elovl6 expression correlated with LD size, which is associated with expression of the LD-associated protein Cidec. Large LDs may obstruct hepatic blood flow by compressing the sinusoids,22 although the direct link between LD size and NASH development and progression is not yet understood.

Inhibiting Elovl6 results in the accumulation of palmitate, which is paradoxically the most potent inducer of cell injury. Our results provide evidence that palmitate is not a bona fide lipotoxic agent and that Elovl6-mediated conversion to stearate or further lipid metabolites is a key step to hepatic lipotoxicity. The underlying mechanism whereby C18 FA species exert more toxic effects than C16 is currently unknown. No difference in the relative amounts of C16:0, C16:1n-7, C18:0, or C18:1n-9 in total hepatic lipids by Elovl6 deficiency after an AHF diet suggests that compositional changes in Elovl6-related FAs in some specific lipids, even though small in quantity, can exert cellular functions important for NASH development. Changes in FA composition observed in both gain and loss of Elovl6 function were very consistent and almost similar among different hepatic lipids (TGs, cholesterol esters, and phospholipids) (Supporting Tables 3-11). Even if the different types of accumulated lipids cause lipotoxicity in the liver additively in different ways, Elovl6 could mediate each event.

It is intriguing that hepatic Elovl6 activity influences the composition of n-6 polyunsaturated fatty acid, elevation of particularly linoleic acid (C18:2n-6), and reduction of dihomo-γ-linolenic acid (C20:3n-6) and arachidonic acid (C20:4n-6) in the total lipid fraction as in TG, cholesterol ester, and phospholipid fractions in Elovl6−/− mice (Fig. 1E and Supporting Tables 3-6) with inverse trends in Ad-Elovl6 overexpression (Fig. 5E and Supporting Table 7-10). C20:4n-6 is the precursor of eicosanoids, especially prostaglandin E2, an arachidonic acid metabolite of the cyclooxygenase pathway as well as leukotriene biosynthesis pathway by arachidonate 5-lipoxygenase, all of which are reported to be involved in NAFLD.23-25 The protective effect of Elovl6 deficiency could not only be a result of a reduction in C18 FAs but also altered n-6 polyunsaturated fatty acid composition, which might have different effects on the inflammatory cascade, ROS generation, kinase activity, and apoptosis. The origin, location, and distribution of these FAs within specific lipid classes require further investigation.

The current data imply the clinical relevance of human NASH. Hepatic expression of lipogenic genes increases in NAFLD patients, and these changes could contribute to increased FA synthesis.26 ELOVL6 expression was consistently increased in the livers of NASH patients, and was positively correlated with steatosis and serum ALT levels. Genetic variations in ELOVL6 are associated with insulin sensitivity.27 Given the important biological activities of many lipids, further analysis on lipidomics of patient liver samples could provide potential insight into the relationship between ELOVL6 and NASH pathophysiology.

The elucidation of the triggering factors for inflammation in NASH is of emerging importance. Increased levels of saturated FAs, particularly palmitate, have been reported in NASH patients9 and have been implicated in NLRP3 inflammasome activation.16, 17 It remains to be determined whether the effect of palmitate on the inflammasome is palmitate itself or occurs via intermediate products of palmitate metabolism. Our findings reveal that FA metabolites through Elovl6, rather than palmitate itself, exert NLRP3 inflammasome activation and IL-1β release. Inflammasome-mediated danger signals released from damaged hepatocytes could trigger inflammasome activation in immune cells, and induce liver inflammation and injury. As previous studies have shown that NADPH oxidase-dependent generation of ROS is essential for NLRP3 inflammasome activation,28 and lipid metabolites including ceramide and diacylglycerol can activate NADPH oxidase and enhance ROS generation,21 our findings indicate that Elovl6 may lead to inflammasome activation through the ROS signaling pathway.

In conclusion, we provide evidence that it is not only the lipid amount, but also the lipid quality and FA composition, that contribute to the fate of NAFLD and pathology of NASH and that Elovl6 is a key player in this process. Understanding the differential effects of specific FAs on the downstream pathway is critical for understanding NASH development and progression. Our findings suggest that Elovl6 is an attractive therapeutic target for NASH and associated complications such as cirrhosis and hepatocellular carcinoma.


We thank Alyssa H. Hasty for critical reading of the manuscript; Katsuko Okubo, Yuko Tamai, and Chizuko Fukui for technical assistance; and members of our laboratories for discussion and comments on the manuscript.