AIM: To investigate the role of key iron-regulatory protein, hepcidin in non-alcoholic fatty liver disease (NAFLD).

METHODS: Hepcidin (Hamp1) knockout and floxed control mice were administered a high fat and high sucrose (HFS) or a regular control diet for 3 or 7 mo. Steatosis, triglycerides, fibrosis, protein and gene expression in mice livers were determined by histological and biochemical techniques, western blotting and real-time polymerase chain reaction.

RESULTS: Knockout mice exhibited hepatic iron accumulation. Despite similar weight gains, HFS feeding induced hepatomegaly in floxed, but not knockout, mice. The livers of floxed mice exhibited higher levels of steatosis, triglycerides and c-Jun N-terminal kinase (JNK) phosphorylation than knockout mice. In contrast, a significant increase in fibrosis was observed in knockout mice livers within 3 mo of HFS administration. The hepatic gene expression levels of sterol regulatory element-binding protein-1c and fat-specific protein-27, but not peroxisome proliferator-activated receptor-alpha or microsomal triglyceride transfer protein, were attenuated in HFS-fed knockout mice. Knockout mice fed with regular diet displayed increased carnitine palmitoyltransferase-1a and phosphoenolpyruvate carboxykinase-1 but decreased glucose-6-phosphatase expression in the liver. In summary, attenuated steatosis correlated with decreased expression of lipogenic and lipid storage genes, and JNK phosphorylation. Deletion of Hamp1 alleles per se modulated hepatic expression of beta-oxidation and gluconeogenic genes.

CONCLUSION: Lack of hepcidin expression inhibits hepatic lipid accumulation and induces early development of fibrosis following high fat intake. Hepcidin and iron may play a role in the regulation of metabolic pathways in the liver, which has implications for NAFLD pathogenesis.

Non-alcoholic fatty liver disease (NAFLD) encompasses a spectrum of liver disease ranging from simple benign steatosis to non-alcoholic steatohepatitis (NASH). NASH, a more aggressive form of disease, is characterized by the presence of lobular inflammation, fibrosis, hepatocellular ballooning and Mallory-Denk bodies[1,2]. NASH with progressive fibrosis can progress to cirrhosis and end stage liver disease[1,3,4].

The precise mechanisms of NASH development are not well understood. Although a so-called “two-hit hypothesis”[5] has been widely adopted[6,7], NASH can also develop in the absence of insulin resistance and simple benign steatosis (i.e., initial hit)[8]. The potential candidates regarded as the “second hit” include oxidative stress, inflammation and changes in mitochondrial function[7,9-12]. Iron is also considered as a “second hit” in liver injury[13] and a role for iron has been reported in NASH pathogenesis. Patients with NAFLD/NASH frequently display elevated serum iron indices and hepatic iron content[14,15]. A strong correlation between hepatic iron content and the level of liver fibrosis in NAFLD/NASH patients has been shown[16-18]. Phlebotomy has also been suggested to alleviate insulin resistance in NAFLD patients[19].

The mechanisms by which iron contributes to NAFLD/NASH pathogenesis have mainly been attributed to oxidative stress, which can induce lipid peroxidation[20] and ultimately the activation of fibrotic signaling[21]. Studies with genetic haemochromatosis (GH) patients have shown the association of primary iron overload with fibrogenesis[22]. By using dietary experimental models, some studies have also suggested a reverse connection between iron and steatosis in rat livers[23,24]. In contrast, another study with a mouse dietary model of iron and high fat failed to show any significant effect of iron on steatosis[25]. The consequences of altered iron homeostasis for lipid metabolism in the liver are therefore unclear.

In this study, we employed hepcidin knockout mice with iron overload phenotype as an experimental model to further study the role of iron metabolism in NAFLD/NASH. Hepcidin is the central regulator of iron homeostasis, which is primarily synthesized in hepatocytes as a circulatory protein[26]. Unlike humans, which express only one hepcidin gene, HAMP, mice express two hepcidin genes, hepcidin (Hamp1) and Hamp2[27]. Hamp1, the human equivalent of mouse hepcidin gene, is by itself sufficient to regulate iron metabolism[28,29]. Hepcidin controls iron homeostasis by decreasing iron absorption from the absorptive enterocytes in the duodenum and the release of iron from the macrophages[30]. The lack of hepcidin expression in knockout mice and in human iron disorders results in iron accumulation both in the liver and other organs[30-32]. GH patients also display impaired hepcidin expression[33]. Although changes in both serum and liver hepcidin expression levels have been reported in NAFLD/NASH patients[14,34-38], the significance of hepcidin in disease pathogenesis is unknown. Our findings in this study with Hamp1 knockout mice administered a high fat diet for different time points suggest a role for hepcidin in NAFLD/NASH pathogenesis. This mouse model may also serve as a novel experimental model of NAFLD/NASH.

To study the interaction of hepcidin-induced iron overload and lipid metabolism, ubiquitous Hamp1 knockout and floxed control mice were administered either high fat and HFS or regular (control) diets, as described in Material and Methods. Since NAFLD/NASH progression can occur over a long period of time, mice were fed up to 7 mo. We have previously shown that the deletion of both Hamp1 alleles induces significant iron overload in the livers of Hamp1 knockout mice by using inductively coupled mass spectrometry (ICP-MS)[29]. ICP-MS analysis did not detect any significant level of iron in the livers of homozygous Hamp1 floxed control mice. Gradual iron deposition was also indicated macroscopically by the darker color of knockout mice livers compared to those of floxed control mice (Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Macroscopic changes in Hamp1 floxed and knockout mice fed with either a high fat and high sucrose or a regular control diet for 3 or 7 mo. Representative images showing the abdominal cavity of mice were obtained with a digital camera (Nikon).

Macroscopic analyses have confirmed that HFS intake induced hepatomegaly and more pronounced visceral fat accumulation in floxed control mice compared to knockout mice (Figure 1). In agreement, the liver weights of floxed mice were significantly higher (3.5 ± 0.46 g) than those of knockout mice (2.42 ± 0.54 g) particularly following 7 mo of HFS administration (Figure 2A and B). However, HFS intake induced similar increases in body weights in both floxed (Figure 2C) and knockout (Figure 2D) mice after either 3 or 7 mo-long feeding, as compared to respective controls fed with regular diet.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Liver and body weights of Hamp1 floxed and knockout mice fed with high fat or regular diets. Average liver (A and B) and body (C and D) weights of floxed (A and C) and knockout (B and D) mice prior to (initial) and after feeding with high fat and sucrose (HFS) or regular control diets for 3 or 7 mo are shown as gram weight.

To further understand these discrepancies between floxed and knockout mice, histological analysis were performed. Hematoxylin and eosin staining of livers showed significantly higher levels of steatosis in floxed than in knockout mice both after 3 and 7 mo-long HFS feeding (Figure 3). The quantification of hepatic triglycerides further confirmed that HFS intake significantly increased hepatic triglyceride content to different extents in floxed and knockout mice (Figure 4). At the end of 3 mo-long high fat intake, the level of hepatic triglyceride accumulation was 2.85-fold higher in floxed mice compared to knockout mice (1876.64 ± 370.84 and 657.98 ± 186.89 μmol/L per 100 g BW) (Figure 4A). Seven mo-long feeding yielded 2.07-fold higher hepatic triglyceride content in floxed than in knockout mice (1837.71 ± 118.12 and 886.91 ± 89.51 μmol/L per 100 g BW) (Figure 4B).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Liver histology in Hamp1 floxed and knockout mice fed high fat or regular diets. Liver sections from floxed and knockout mice fed with high fat and sucrose (HFS) or regular diets for 3 and 7 mo were stained with hemotoxylin and eosin. Representative images obtained with a Nikon Eclipse E400 light microscope are shown (20 ×). Arrows indicate steatosis.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Liver triglyceride content in Hamp1 floxed and knockout mice fed high fat or regular diets. Hepatic triglyceride content in floxed and knockout mice fed with regular or high fat sucrose (HFS) diets for 3 (A) or 7 (B) mo was quantified using 50 mg of wet liver tissue. Liver triglyceride amount was expressed as μmol per liver per 100 g body weight (μmol/L per 100 g BW).

Sirius Red staining of liver sections showed that knockout, but not floxed, mice developed fibrosis within 3 mo of high fat intake (Figure 5A). The deletion of both Hamp1 alleles per se has also caused weaker but significant level of fibrosis in the livers of knockout mice (Figure 5A). Quantification by ImageJ analysis has shown a 2.56-fold higher level of fibrosis in the livers of high fat-fed knockout than regular diet-fed knockout mice at 3 mo (Figure 5B) In contrast to 3 mo, 7 mo of high fat intake induced fibrosis in the livers of floxed mice (Figure 6A). Compared to 3 mo, regular diet feeding for 7 mo slightly increased the level of fibrosis in knockout mice livers (Figure 6A). Knockout mice with 7 mo-long high fat intake developed the highest level of fibrosis, as shown by Image J quantification (Figure 6B). The hepatic expression patterns of alpha smooth muscle actin (αSMA) protein, a marker for hepatic stellate cell activation, were in agreement with our histological analysis. Three months-long HFS feeding elevated liver αSMA expression in knockout, but not floxed, mice, as shown by Western blotting (Figure 7A). The deletion of Hamp1 alleles by itself increased hepatic αSMA expression (Figure 7A). In contrast to 3 mo, 7 mo-long high fat intake increased αSMA expression in the livers of both floxed and knockout mice (Figure 7A).

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Fibrosis in Hamp1 floxed or knockout mice fed high fat or regular diets for 3 mo. A: Fibrosis in the livers of floxed and knockout mice fed on regular or high fat sucrose (HFS) diets for 3 mo was detected by Sirius Red staining of tissue sections. Representative images obtained with Nikon Eclipse E400 light microscope are shown; B: 10 independent images (10 x) taken from each group were quantified using ImageJ ROI manager software. The collagen proportional area (CPA) was determined by calculating the percentage of collagen-occupied pixels against the total pixel values.

Open in New Tab   Full Size Figure   Download Figure

Figure 6 Fibrosis in Hamp1 floxed or knockout mice fed high fat or regular diets for 7 mo. Liver fibrosis in floxed and knockout mice fed on regular or high fat sucrose (HFS) diets for 7 mo was detected (A) and quantified (B), as described above. CPA: Collagen proportional area.

Open in New Tab   Full Size Figure   Download Figure

Figure 7 Protein expression levels of phosphorylated Jun N-terminal kinase and alpha smooth muscle actin in Hamp1 floxed and knockout mice fed with high fat or regular diets for 3 or 7 mo. The expression levels of alpha smooth muscle actin (αSMA) (A) and phosphorylated Jun N-terminal kinase (p-JNK) (B) proteins in the livers of floxed and knockout mice fed with regular or high fat sucrose (HFS) diets for 3 or 7 mo was determined by western blotting, as described in Material and Methods. An anti-gapdh antibody was used as control to determine equal protein loading; Gapdh: Glyceraldehyde 3-phosphate dehydrogenase.

Studies with JNK knockout mice fed with methionine-choline-deficient diet (MCD) diets have indicated a role for c-Jun N-terminal kinase (JNK) in steatosis[42]. JNK is activated by phosphorylation on serine residues[43]. The expression levels of phosphorylated JNK protein in the livers of Hamp1 transgenic mice were therefore determined by western blotting using specific anti-phospho JNK antibodies (Figure 7B). Three-month-long high fat intake significantly stimulated JNK phosphorylation in the livers of floxed, but not knockout, mice (Figure 7B). In contrast, the effect of high fat intake on JNK phosphorylation in the liver was weakened by 7 mo-long feeding (Figure 7B).

To further investigate the underlying mechanisms of attenuated fat accumulation in the livers of knockout mice with high fat intake, mRNA expression levels of genes, which are known to be involved in lipid metabolism, were examined by real-time PCR (Figure 8). The transcription factor, sterol regulatory element-binding protein-1c (Srebp-1c) is involved in de novo lipogenesis and its expression is also regulated at the transcriptional level[44,45]. The deletion of Hamp1 alleles did not significantly alter basal hepatic expression levels of Srebp-1c (Figure 8A and B). Three months of high fat intake stimulated Srebp-1c expression by 13.39-fold in floxed and 7.40-fold knockout mice compared to controls (Figure 8A). In contrast, 7 mo of high fat intake elevated Srebp-1c expression only by 3.72-fold in floxed mice (Figure 8B). Furthermore, 7 mo-long high fat feeding did not significantly alter liver Srebp-1c expression in knockout mice (Figure 8B).

Open in New Tab   Full Size Figure   Download Figure

Figure 8 Expression of genes involved in lipogenesis, lipid storage and secretion. The mRNA expression levels of Srebp-1c (A and B), Fsp27 (C and D), and Mttp (E and F) in the livers of floxed and knockout mice fed with regular and high fat sucrose (HFS) diets, was determined by real-time polymerase chain reaction. Gene expression in high fat-fed floxed or knockout and regular diet-fed knockout mice for 3 (A, C and E) or 7 mo (B, D and F) was expressed as fold change of that in floxed mice fed with a regular diet for the same time period.

Fat-specific protein-27 (Fsp27) protein is involved in lipid droplet formation[46]. HFS feeding for 3 and 7 mo significantly induced Fsp27 expression in the livers of floxed mice by 3.83- and 5.36-fold, respectively compared to regular diet-fed floxed mice (Figure 8C and D). The livers of knockout mice fed with HFS for 3 or 7 mo displayed significantly lower induction of Fsp27 expression than floxed mice, which was more prominent at 7 mo (Figure 8C and D). Liver Fsp27 expression was not significantly altered in knockout mice fed with regular diets for 3 or 7 mo compared to respective floxed controls (Figure 8C and D).

Microsomal triglyceride transfer protein (Mttp) protein is responsible for the production and secretion of VLDL particles[47]. The mRNA expression level of Mttp in the liver was not significantly altered in floxed and knockout mice after 3 mo of high fat intake (Figure 8E). However, high fat exposure for 7 mo significantly suppressed Mttp expression in the livers of both floxed and knockout mice (Figure 8F).

Changes in fatty acid oxidation in the liver play an important role in NAFLD pathogenesis. Peroxisome proliferator-activated receptor-alpha (Pparα) activates the transcription of genes involved in the regulation of fatty acid β-oxidation[48]. The mRNA expression levels of Pparα were up-regulated at similar levels in the livers of both floxed and knockout mice within 3 mo of high fat feeding (Figure 9A). In contrast, the livers of floxed and knockout mice with 7 mo of high fat exposure displayed significantly inhibited Pparα expression (Figure 9B). Carnitine palmitoyltransferase-1 (Cpt1) is the rate-limiting enzyme in mitochondrial β-oxidation pathway[49]. Three month-long high fat administration did not exert a significant effect on hepatic Cpt1a expression in floxed and knockout mice (Figure 9C). On the other hand, the livers of knockout mice fed with regular diet for 7 mo expressed higher Cpt1a levels compared to floxed mice fed under similar conditions, suggesting a role for gradual iron deposition (Figure 9D). Seven month-long high fat intake did not alter hepatic Cpt1a expression in floxed mice (Figure 9D). In contrast, long-term high fat exposure significantly suppressed Cpt1a expression in the livers of knockout mice compared to knockout controls (Figure 9D).

Open in New Tab   Full Size Figure   Download Figure

Figure 9 Expression of genes involved in β-oxidation and gluconeogenesis. The mRNA expression levels of Pparα (A and B), Cpt1a (C and D), Pck1 (E and F) and G6pc (G and H), in the livers of Hamp1 floxed and knockout mice fed with regular and high fat sucrose (HFS) diets, was determined by real-time polymerase chain reaction. Gene expression in high fat-fed floxed or knockout and regular diet-fed knockout mice for 3 (A, C, E and G) or 7 mo (B, D, F and H) was expressed as fold change of that in floxed mice fed with a regular diet for the same time period.

Both phosphoenolpyruvate carboxykinase-1 (Pck1) and glucose-6-phosphatase (G6pc) are involved in gluconeogenesis. Similar to Cpt1a, the deletion of Hamp1 alleles significantly up-regulated basal Pck1 mRNA expression in the liver. In contrast, the absence of hepcidin expression suppressed basal hepatic G6pc mRNA expression (Figure 9E-H). Both 3 and 7 mo-long high fat exposure significantly inhibited Pck1 and G6pc mRNA expression in the livers of both floxed and knockout mice (Figure 9E-H).

Changes in iron metabolism contribute to liver injury[22,50]. The deposition of iron in the liver correlates with disease severity in NAFLD patients[15]. The mechanisms by which excess iron contribute to NAFLD pathogenesis is unclear. Although inconclusive, some studies suggested a role for iron in the regulation of lipid metabolism[23-25]. Since hepcidin is the central regulator of iron metabolism, we investigated its role in fatty liver disease. We and others showed iron accumulation in Hamp1 knockout mice[29,31,51]. Hamp1 knockout mice were administered high fat diets for different time periods to generate pathological features in the liver, which are representative of NAFLD/NASH[2]. Collectively, our findings showed a strong correlation between hepcidin and lipid metabolism, and fibrosis in the liver.

The absence of hepcidin expression in Hamp1 knockout mice exerted an inhibitory effect on hepatic lipid accumulation. This effect was not due to altered rates of diet consumption or weight gain and suggests the involvement of regulatory mechanisms. Previous studies showed a converse relationship between iron and lipid metabolism[22,23]. Since lack of hepcidin expression causes iron overload, elevated hepatic iron content may have interfered with fat accumulation in HFS-fed knockout mice. Furthermore, our findings suggest a role for JNK in this process. Namely, we showed a direct correlation between JNK phosphorylation and steatosis levels in floxed mice livers. In contrast, the livers of Hamp1 knockout mice did not display significant JNK phosphorylation. Of note, the deletion of JNK1 reverses steatosis[52,53] and JNK is activated by phosphorylation[43]. Hepcidin-mediated changes in JNK activation may therefore be associated with attenuated steatosis in Hamp1 knockout mice, particularly in early stages of high fat exposure.

Besides iron and JNK, altered metabolic gene expression in high fat-fed knockout mice may play a role in the inhibition of lipid accumulation. This is supported by our findings, which showed that the hepatic expression level of genes involved in lipogenesis and lipid storage do not adequately respond to high fat intake in Hamp1 knockout mice. Namely, Srebp-1c and Fsp27 expression were blunted in the livers of HFS-fed knockout, but not floxed, mice. These findings are significant because Srebp-1c and Fsp27 expression are regulated at mRNA level[54]. Furthermore, the deletion of Hamp1 alleles did not alter their basic expression levels. Iron-deficient rodents have been reported to display elevated lipogenic gene expression, which indirectly supports our findings[55-57]. Hepatic lipid homeostasis is also regulated by lipid export via VLDL secretion. The hepatic expression levels of Mttp, which is important in this process, were comparable between control and knockout mice. Our findings therefore suggest that decreased lipogenesis and lipid storage, but not increased lipid secretion, might lead to attenuated steatosis in high fat-fed Hamp1 knockout mice.

Increased mitochondrial β-oxidation alleviates extra-hepatic fat burden in NAFLD by disposing of excess lipids[58]. Pparα, which induces the transcription of genes involved in β-oxidation, is itself regulated at the transcriptional level[59,60]. However, Pparα is not expected to contribute to liver pathology in Hamp1 knockout mice because HFS-fed floxed and knockout mice livers displayed similar levels of Pparα expression. Cpt1 is the rate-limiting enzyme in β-oxidation. Long-term high fat intake significantly suppressed Cpt1a expression only in knockout mice livers suggesting a role for it in attenuated steatosis in Hamp1 knockout mice. Interestingly, Hamp1 deletion by itself elevated hepatic Cpt1a expression. Besides β-oxidation, mitochondria is also important for iron metabolism[61]. It is feasible that iron accumulation caused by Hamp1 deletion modulates metabolic gene expression in mitochondria. Of note, mitochondrial changes contribute to NAFLD/NASH pathology[11]. Hamp1 deletion also altered the expression of gluconeogenic genes, Pck1 and G6pc. Hepcidin serves as a gluconeogenic sensor in mice during starvation[62]. The reasons for the differential regulation of Pck1 and G6pc expression in knockout mice livers are unclear. Pck1 and G6pc are however regulated by various transcription factors including Foxo1[54] and iron regulates Foxo1 in adipocytes[63]. The net effect of hepcidin and iron on metabolic processes in the liver requires further investigation.

Despite amelioration of steatosis, high fat administration caused injury in the livers of Hamp1 knockout mice. In fact, knockout mice displayed an earlier and more pronounced development of fibrosis compared to control mice. Previous studies using MCD experimental models have shown that iron supplementation attenuates steatosis and triggers fibrosis[24,64]. Of note, MCD diet does not reproduce the metabolic changes observed in NAFLD/NASH patients and induces weight loss[65,66]. On the other hand, most high fat diet models induce metabolic changes but not fibrosis[66,67]. Furthermore, introduction of iron in the diet can create secondary effects by up-regulating liver hepcidin synthesis and thereby inhibiting the expression of iron exporter, ferroportin[68-70]. This will then lead to sequestration of iron in Kupffer cells and trigger inflammation. These artefacts are avoided in in our experimental system because iron accumulation is directly caused by the lack of hepcidin expression. Our high fat-fed Hamp1 knockout mice, which develop early fibrosis, may therefore be an advantageous NAFLD/NASH model.

Simple steatosis is considered to be a benign condition in NAFLD patients. In vivo and in vitro studies have also shown this to be a beneficial process because triglycerides synthesis protects the liver from lipotoxicity induced by free fatty acid accumulation[64,71]. The decreased level of steatosis in synergy with iron might be responsible for early fibrosis development in the livers of HFS-fed Hamp1 knockout mice.

In summary, our findings strongly suggest a role for hepcidin in the regulation of hepatic lipid and carbohydrate metabolism. There are currently a limited number of NASH experimental models[66]. Hamp1 knockout mice will therefore be useful to investigate the molecular mechanisms of metabolic processes and fibrosis in NASH pathogenesis.

Lack of hepcidin expression due to the deletion of Hamp1 alleles inhibited lipid accumulation in the liver following a high fat and high sucrose diet administration. Lack of c-jun kinase phosphorylation and the changes in the expression of metabolic genes, which are involved in lipogenesis and lipid storage, played a role in attenuated steatosis observed in hepcidin knockout mice. Knockout mice developed fibrosis within 3 mo of high fat exposure, which was more prominent at 7 mo. Deletion of Hamp1 alleles by itself modulated hepatic expression of genes involved in mitochondrial fatty acid oxidation and gluconeogenesis. In summary, hepcidin is associated with the regulation of metabolic processes in the liver and the lack of hepcidin expression triggers early fibrosis development. High fat-fed hepcidin knockout mice may therefore serve as a useful animal model to study different aspects of fatty liver disease pathogenesis.