©2012 Baishideng Publishing Group Co., Limited. All rights reserved.
World J Gastroenterol. Sep 14, 2012; 18(34): 4651-4658
Published online Sep 14, 2012. doi: 10.3748/wjg.v18.i34.4651
Interactions between hepatic iron and lipid metabolism with possible relevance to steatohepatitis
Umbreen Ahmed, Department of Physiology, Army Medical College, National University of Sciences and Technology, Rawalpindi 46000, Pakistan
Patricia S Latham, Department of Pathology, The George Washington School of Medicine and Health Sciences, Washington, DC 20037, United States
Phillip S Oates, Physiology M311, Anatomy, Physiology and Human Biology, University of Western Australia, Perth, WA 6009, Australia
Author contributions: All authors contributed equally to the production of this manuscript.
Correspondence to: Dr. Phillip S Oates, Physiology M311, Anatomy, Physiology and Human Biology, University of Western Australia, Perth, WA 6009, Australia. email@example.com
Telephone: +61-8-64881391 Fax: +61-8-64881025
Received: January 10, 2012
Revised: March 27, 2012
Accepted: March 29, 2012
Published online: September 14, 2012
ROLE OF IRON IN HEPATIC LIPID METABOLISM: A BRIEF OVERVIEW
The liver is a major site for the storage of iron and the metabolism of lipids and is therefore an important site for interaction between these two metabolic pathways. One role of iron in the pathogenesis of diseases associated with hyperlipidemia and lipid deposition is likely to include an ability to induce oxidative stress and inflammation in the liver. In the setting of lipid deposition in the liver, iron overload has been associated with increased fibrosis and progression of non-alcoholic fatty liver disease and non-alcoholic steatohepatitis (NAFLD/NASH)[1-3]. Iron in its ferrous form can generate free radicals via the Fenton reaction, resulting in oxidative stress and lipid peroxidation. These effects can, in turn, stimulate hepatic stellate cells to increase production of collagen and progression of fibrosis in patients with NAFLD and NASH. Notwithstanding these indirect effects, there are also recent and archival studies to show that iron has direct effects on lipid metabolism. Variations in hepatic iron stores result in inappropriate lipogenesis for lipid storage and secretion. For example, iron deficiency is associated with increased hepatic lipogenesis and lipemia[4,5]. Since the liver co-ordinates iron and lipid metabolism, and under conditions of stress is the site of excessive deposition of both these nutrients, this review will focus on the role of iron in the metabolism of hepatic lipids. In particular, it examines the role of different enzymes, receptors and transporters, as well as the effects of proteins involved in lipogenesis and lipoprotein secretion as sites for interaction between iron and lipids. The review then addresses the way in which iron may function in the progression of NAFLD. To begin, a brief overview of normal hepatic lipid metabolism is provided.
HEPATIC LIPID METABOLISM
Hepatic steatosis occurs whenever there is an imbalance among the uptake, synthesis, oxidative and secretory pathways of fatty acid metabolism. Sources of fatty acids for hepatic triglyceride (TG) synthesis are those derived from the plasma non-esterified fatty acids (NEFA) pool and those from within the liver through de novo lipogenesis (DNL), mainly from glucose (Figure 1). The hepatic uptake of fatty acid from the plasma is not regulated and is therefore directly related to the concentration of plasma NEFAs. Adipose tissue by its TG lipolysis is the main contributor to the plasma NEFA pool. The remaining NEFAs come from peripheral lipolysis of lipoproteins [chylomicrons (CM), very low density lipoprotein (VLDL)] and dietary fatty acids in the circulation. The hepatic uptake of fatty acids occurs by fatty acid binding protein and fatty acid translocase. Once in the hepatocyte, fatty acids undergo oxidation or they may be detoxified by re-esterification with glycerol and cholesterol to form TG and cholesteryl esters (CE), respectively. The TG and CE are then secreted as VLDL into the plasma or, if in excess, are stored in the cytoplasm as lipid droplets. Adipose differentiation-related protein (ADRP) is expressed in hepatocytes and is involved in cytosolic lipid storage.
Figure 1 Possible sites (shown in bold) of interactions between iron and lipid metabolism.
Iron deficiency can have its effects by generating anaemia/hypoxia while iron overload can interact by generating oxidative stress/lipid peroxidation and cytokine production. The grey area shows the basolateral surface of the hepatocyte bathed in the sinusoidal fluid. The green area is the canalicular surface through which bile is actively secreted into the biliary canaliculi. The orange area shows the lipoproteins and enzymes acting in peripheral circulation. CM: Chylomicron; CMR: Chylomicron remnant; VLDL: Very low density lipoprotein; IDL: Intermediate density lipoprotein; LDL: Low density lipoprotein; HDL: High density lipoprotein; LDLR: LDL (apo B/E) receptor; LRP: LDL receptor related protein; FABP: Fatty acid binding protein; FAT/CD36: Fatty acid translocase; SR-B1: Scavenger receptor-B1; NEFA: Non-esterified fatty acid; CE: Cholesteryl ester; SFA: Saturated fatty acid; MUFA: Mono-unsaturated fatty acid; TG: Triglycerides; apo-B: Apolipoprotein-B; DNL: de novo lipogenesis; LD: Lipid droplet; LPL: Lipoprotein lipase; HL: Hepatic lipase; LCAT: Lecithin cholesterol acyltransferase; GPAT: Glycerol-3-phosphate acyltransferase; DGAT: Diacylglycerol acyltransferase; SCD: Steroyl CoA desaturase; FAS: Fatty acid synthase; HMG CoA R: HMG CoA reductase; 7αOHase: 7α hydroxylase; ACAT: Acyl-CoA cholesterol acyltransferase; MTP: Mitochondrial Trifunctional protein; ADRP: Adipose differentiation-related protein.
The partitioning of hepatic TG into different pools under normal conditions of fasting/feeding and dietary variations, i.e. carbohydrate or fat diet, involves different hormonal and metabolic regulatory mechanisms. Insulin is a fat-sparing hormone and a major metabolic factor in the regulation of fat metabolism. In the liver, insulin increases TG synthesis by the hepatocytes and restricts its secretion by partitioning the newly synthesized TG into cytosolic stores. In addition, insulin also inhibits lipolysis in the hepatocyte, decreases the assembly of VLDL by impairing the association of apoB-100 with TG, inhibits the translation of apoB-100 mRNA in human hepatocellular carcinoma (HepG2) cells and stimulates the degradation of newly synthesized apoB, which collectively decrease VLDL-TG secretion. These actions coupled with insulin-stimulated expression of ADRP promotes lipid droplet assembly. Furthermore, insulin also activates sterol regulatory element binding protein-1c, which transcriptionally activates nearly all genes involved in DNL.
EFFECTS OF IRON DEFICIENCY ON HEPATIC LIPIDS
There is evidence to suggest that lipoprotein uptake pathways are affected by iron.
Anaemia/hypoxia is associated with decreased fat uptake by the liver, but increased fat accumulation by the liver. Decreased fat uptake may either be due to impaired peripheral lipolysis or decreased uptake of CM and VLDL remnants, leaving the lipoproteins to accumulate in the circulation. Rat models of chronic anemia/hypoxia in which iron deficiency is present show decreased lipoprotein lipase and hepatic lipase activities[14-18]. This may be due to decreased synthesis, defects in the structure or conversion from precursor to the functional form, or in the activation of these enzymes[14,16]. It may be that iron itself can impact on hepatic uptake pathways. Supporting this idea, Brown et al have shown that the binding of low density lipoprotein to its receptors requires a divalent cation. Although speculative, one of the divalent cations may include ferrous-iron. Anaemic rats produced by experimentally induced chronic renal failure show down- regulation of the expression of the VLDL-receptor in a non-erythropoietin dependent manner. Furthermore, the plasma ratio of cholesteryl ester to cholesterol is increased with iron deficiency in rats. This cannot be explained by increased lecithin cholesterol acyl transferase (LCAT) activity, since LCAT is reduced in iron deficiency[21,22].
On the other hand, iron deficiency in rats has been shown to increase hepatic lipogenesis, leading to cellular TG accumulation and steatosis[4,5,23,24]. Given that hepatic uptake pathways are down regulated, the increased lipogenesis could be due to decreased utilization of fatty acids or increased DNL. In iron deficient rats, it is proposed that beta-oxidation of fatty acids is decreased, allowing the fatty acids to be diverted towards TG synthesis. Supporting this, the level of hepatic carnitine, the long chain fatty acid transporter required for β-oxidation of fatty acids is reduced in the iron deficient state. Since this enzyme requires iron as a cofactor, its activity might be expected to decrease with iron deficiency[24,25]. However, others report that beta-oxidation of fatty acids remains unchanged, while hepatic DNL increases[4,23]. The increased hepatic lipogenesis seen in iron-deficient rats is supported by genomic studies that show a concurrent decrease in gene expression related to hepatic β-oxidation and an increase in gene expression related to lipogenesis. Therefore, some of the changes in lipid metabolism appear to be related to effects of iron-deficiency on transcriptional/post-transcriptional mechanisms, as well as to effects on the kinetics/activity of enzymes that depend upon iron as a cofactor.
In contrast to the studies that show an increase in hepatic lipogenesis with iron deficiency, others have reported a decrease in hepatic lipid synthesis. In rats with moderate iron deficiency, glucose-fueled hepatic lipogenesis fell which was attributed to decreased activity of either the rate limiting enzymes [fatty acid synthase (FAS)] or enzymes involved in providing the redox potential for lipogenesis (glucose 6 phosphate dehydrogenase, malic enzyme)[28-30].
In chicks, anemia depresses levels of mitochondrial cytochromes[31-33]. This finding is consistent with that found in yeast in which iron-deficiency down-regulates transcription of certain respiratory cytochromes that require heme as prosthetic groups for activity, while the heme degradation enzyme, heme oxygenase, is upregulated. It is proposed that this decrease in the activity of hepatic heme pathways in iron-deficiency preserves the availability of iron for other essential sites/pathways when cellular iron levels are low. Supporting this idea is the observation that when lipid metabolism is affected by iron deficiency in yeast, amino acid homeostasis remains unaltered.
Iron deficiency can affect the function of many hepatic enzymes involved in cholesterol metabolism. In a mouse model of iron-deficiency induced by genetic mutation of hephaestin (sla mice), lack of the copper-dependent ferroxidase in the enterocyte results in impaired iron absorption and anaemia. Under these conditions, hepatic cholesterol synthesis is reduced in association with decreased mevalonic acid levels. HMG CoA reductase activity, the rate limiting enzyme in de novo cholesterol synthesis, is increased in this model, arguing that this enzyme is not responsible for the decreased cholesterol production. Similarly, acetyl-CoA thiolase and HMG-CoA synthase involved in cholesterol production are not affected by iron deficiency. All these enzymes act upstream in the production of mevalonic acid, suggesting that this intermediate is shunted to a non-steroidal pathway.
Other studies show that the cholesterol degradation pathway is also affected. One study in dogs shows that iron deficiency can decrease the activity of hepatic 7α hydroxylase, the rate limiting enzyme in the conversion of cholesterol into bile acids. In this study, bile contained cholesterol crystals, suggesting that reduced conversion of cholesterol to bile acids may impair biliary micellar formation and explain the cholesterol precipitates in the bile. However, hepatic lipids were not measured in this study.
Nutritional iron depletion in rats also leads to alterations in the fatty acid composition of hepatic phospholipids, indicating impaired desaturation of saturated and essential fatty acids[29,39,40]. Steroyl CoA desaturase (SCD) is required for the conversion of saturated fatty acids to the unsaturated form and this enzyme is dependent on iron for its activity. Lower activity of SCD in iron deficiency may explain the increased saturated composition of fatty acids in the liver in this condition. The effect of iron on the various steps in lipid metabolism is summarised in Figure 1.
It is likely that some of the reasons for the conflicting observations reported here as a result of iron deficiency can be explained in part by the specifics of the model system, by the means of inducing iron deficiency through phlebotomy or dietary restriction, or by the extent of the iron deficiency. Severe iron deficiency will produce anaemia and hypoxia which are likely to activate additional transcription factors to those stimulated during mild iron deficiency, and in turn, affect the transcription of more genes encoding enzymes/proteins. Analysis of the reports cited in this review for correlations between the extent of iron deficiency and alterations in lipid metabolism was not possible, since the level of iron deficiency was not measured in these studies. It is recommended that future studies exploring effects of iron deficiency need to incorporate measurements of iron into their experimental design.
Another reason for the conflicting observations reported here might be the model systems themselves. Some animal model systems are designed to develop iron-deficiency from birth and others induce iron deficiency in a mature animal[4,23]. There are studies to suggest that metabolic conditions in development can impact on adult metabolism. For example, epidemiological data show that under normal birth weight conditions, there is an inverse correlation between the weight at birth and adult obesity[42,43]. Experimental data show that changing the diets during pregnancy or during the newborn period may permanently change the rate of lipid metabolism. Therefore, it is important to recognize the possibility of fetal programming on hepatic lipid metabolism as an explanation for some of the differences reported.
EFFECTS OF IRON OVERLOAD ON HEPATIC LIPIDS
Iron overload can generate oxidative stress[45,46] and lipid peroxidation, which can modify the fatty acid profile of cellular membranes, leading to their disruption, damage to cell organelles[47-49] and impairment of mitochondrial oxidative metabolism[50,51]. It is suggested that the free radicals that form may cause a change in the ratio of saturated to unsaturated membrane phospholipids, leading to alterations in membrane fluidity. This in turn may affect the activity of the embedded enzymes[53-56]. Enzymes respond to oxidative stress differently by altering their activity. It is also known that the formation of peroxidation products by hepatocytes increases with the proportion of unsaturated fatty acids. Thus, polyunsaturated fatty acids in the presence of iron overload may exert an inhibitory effect on lipogenic genes (e.g., FAS) by generating a peroxidative cytotoxic effect .
Iron overload also has direct effects on hepatic lipid metabolism, although studies report conflicting results based on different experimental models. Specific enzymes in hepatic cholesterol metabolism show variable responses to iron overload. In rats with dietary iron overload, an increase in the activity of acyl-CoA cholesterol acyltransferase (ACAT) coupled with reduced activities of HMG CoA reductase and 7 α-hydroxylase correlates with hypercholesterolemia and unaltered hepatic cholesterol content. Since ACAT increases intrahepatic cholesterol esterification and contributes to VLDL-cholesterol secretion, these findings suggest that the cholesterol synthetic and excretory pathways are not affected while the secretory pathways maybe upregulated in this model system[55,58]. On the other hand, a recent study in mice by Graham et al, shows that transcripts of seven enzymes, including the rate limiting enzyme HMG CoA reductase, are up-regulated with increasing hepatic iron, suggesting that hepatic iron loading increases liver cholesterol synthesis. Although valuable, genomic analysis does not provide insight into enzyme kinetics and substrate concentration which may also be altered by the level of iron. This may explain some of the differences between these studies as well as different feeding regimes employed. Hepatic iron overload using carbonyl iron in the methionine-choline deficient rat model of NAFLD is associated with decreased hepatic TG and decreased hepatic steatosis; however, an improvement in liver injury is not seen due to increased necroinflammation and a trend towards increased perivenular fibrosis in iron loaded animals. In carbonyl-iron and iron dextran models of iron overload in mice, studies show an upregulation of the mRNA and enzyme activity of SCD. SCD increases the biosynthesis of unsaturated fatty acids at 8 but not 2 mo of feeding. In contrast, feeding ferric citrate for 3 mo has no effect on hepatic fatty acid composition or desaturase activity. These findings suggest that the experimental model used, the level and duration of iron overload, as well as cellular repartition of iron storage in the liver may play a role in SCD induction. The effect of iron on the various steps in lipid metabolism is summarised in Figure 1.
In vitro studies in HepG2 cells show that iron overload increases intracellular lipid droplet formation by increasing the expression of cluster of differentiation 1d, an unconventional major histocompatibility complex class 1 molecule reported to monitor intracellular and plasma membrane lipid metabolism. Both of these changes are also associated with increased expression of phosphatidylserine in the outer leaflet of the plasma membrane, a feature associated with apoptosis and cell death[62,63]. Hepatosteatosis has also been associated with an increased likelihood of apoptosis. Cell death of mature hepatocytes disproportionate to their ability to regenerate has been referred to as a “third hit” in the pathogenesis of NAFLD/NASH leading to progression of fibrosis[65,66].
Several important factors may contribute to the inconsistencies described here in studies exploring effects of iron overload. These include differences in experimental designs, diets, age, sex, weight and strain of animals; differences in methods for generating iron overload; and also maternal status and its effect on fetal programming, as discussed for models of iron deficiency above.
ROLE OF IRON-OVERLOAD IN NON-ALCOHOLIC STEATOHEPATITIS
NAFLD is now accepted as an hepatic component of the metabolic syndrome. It represents a spectrum of disease ranging from a benign state of non-alcoholic fatty liver to severe fibrosis or cirrhosis. NASH is the intermediate progressive stage between the two conditions. The exact mechanism for this progression is not known but a “two hit theory” is generally accepted. The first hit is considered to be insulin resistance, which leads to excessive accumulation of fat in the liver. The second hit is the generation of oxidative stress by a number of factors, including hepatic iron overload, which lead to cytotoxicity and necroinflammation. A number of clinical and animal studies have shown variable results concerning the effects of iron in NAFLD and NASH. This has lead to controversy about its role in the disease progression (Figure 2).
Figure 2 Schematic diagram depicting disease progression in non-alcoholic fatty liver disease modified after Gentile and Pagliassotti.
Iron overload is generally hypothesised to play a role in generation of second hit. In non-alcoholic fatty liver (NAFL) disease, iron overload in the liver may promote the development and progression of steatohepatitis, working in a feed-forward manner to promote the “third hit”, on-going injury and death of hepatocytes leading to fibrosis and cirrhosis. NASH: Non-alcoholic steatohepatitis.
As already mentioned, excess iron is directly toxic to cells[47-49]. Furthermore, iron accumulation has a proinflammatory and profibrogenic role by activating Kupffer cells to release inflammatory cytokines[68,69] and by activating hepatic stellate cells, which can culminate in the replacement of parenchymal tissue with connective tissue. Bacon et al, in 1994 first showed abnormal iron parameters (serum ferritin and transferrin saturation) and increased hepatic iron concentration in cases of NASH. Other clinical studies have also shown an increase of hepatic iron deposition in NASH/NAFLD and an increased prevalence of the hemochromatosis gene (HFE) mutation[2,3,71-73]. One study showed an association between increased hepatic iron and severity of fibrosis in NASH. In contrast to the work of Bacon and others, however, there are also clinical studies that fail to find an increase in hepatic iron deposition in NASH[74-78] or an increase in HFE mutation[79-81], casting doubt on iron as a causative agent in the progression of the disease.
On the other hand, several clinical studies of NAFLD/NASH show an improvement in liver enzymes and insulin sensitivity with phlebotomies, even in cases without increased levels of iron[82-84]. One case report of a patient with NASH shows complete histologic resolution of steatohepatitis in the liver with phlebotomy-induced iron depletion. However, it must be kept in mind that repeated phlebotomies might predispose to anemia, which could lead to tissue hypoxia and subsequent peripheral vasodilatation. Cardiac output would be expected to increase in response to reduced peripheral resistance, which in turn would increase tissue perfusion and peripheral lipoprotein clearance, reducing the lipid load on the liver. This proposed sequence of events is supported by hemodynamic studies in the anaemic state[86-88] and clearance studies using adrenergic agonists[89,90]. Taken together, an improvement in NAFLD/NASH observed with repeated phlebotomies might be more a physiological response to anaemia/tissue hypoxia rather than a direct effect of reducing iron toxicity. Care must be taken to choose this procedure in cardiac compromised patients.
It is a generally accepted hypothesis that iron overload may play a role in the progression of NAFLD via promoting the 2nd hit (Figure 2). However, it is also possible that the steatosis, cytokines and oxidative stress that are integral to NASH pathophysiology secondarily lead to iron overload, which then works in a feed-forward manner to generate more oxidative stress and inflammation, making the environment more favourable for the 3rd hit, injury that overwhelms the capacity of dying hepatocytes to regenerate and increases scar, as proposed by Diehl[65,91]. Supporting the idea, increased hepatic iron deposition is also seen in non-biliary causes of cirrhosis, such as alcoholic liver disease and chronic viral hepatitis as well as NASH. The process of hepatocellular injury may itself result in increased iron uptake from the intestine. The expression of transferrin receptor might be increased in regenerating hepatocytes after liver injury, which could contribute to increased hepatic iron loading. Another mechanism of increased iron accumulation in liver in NAFLD and NASH might be due to changes in levels of ferroportin and the iron regulatory peptide hepcidin[95,96]. Rats developing insulin resistance on a high fat, high energy diet show decreased hepatic mRNA expression of hepcidin. Similarly, mice developing steatosis on a choline deficient diet show a negative correlation of hepcidin and ferroportin (iron export molecule) mRNA with hepatic lipid concentration, suggesting that enhanced dietary intake and reduced hepatic iron efflux may lead to increased hepatic iron content. In patients with NAFLD, there is also reduced expression of ferroportin and the iron-sensing molecule, hemojuvelin, which is associated with an increase of tumor necrosis factor-α, hepatic iron accumulation and increased hepcidin expression. Hepcidin expression and iron regulatory molecules need to be further investigated in the livers of patients with metabolic syndrome and NAFLD.
It may also be the case that factors inducing localized iron deposition in the microenvironment of the liver are as, or more, important than total hepatic iron or systemic levels of iron. Iron in the microenvironment can promote regional oxidative stress that can lead to changes in microcirculation/capillarization, inflammation and fibrosis in NAFLD. In a rabbit model of steatohepatitis, Otogawa et al have shown an increased phagocytosis of erythrocytes by Kupffer cells that is associated with increased hepatic iron deposition, suggesting that iron may accumulate locally as part of heme degradation. This finding is supported by a clinical study of patients with NASH in which aggregations of erythrocytes are seen in inflammed hepatic sinusoids.
In the aggregate, the findings reviewed here suggest the hypothesis that hepatic iron overload itself may be a critical event in the progression of NAFLD/NASH by exacerbating the 2nd hit of inflammation or perhaps by providing a critical 3rd hit that can ultimately lead to liver cell loss, failure of hepatocyte replication and eventual cirrhosis. Further studies are required to clearly establish the relationship between NASH/NAFLD, insulin resistance and iron overload.
Peer reviewers: Dr. Tamir Miloh, Phoenix Children’s Hospital, 1919 E Thomas Rd, Phoenix, AZ 85016, United States; Manuel Romero-Gomez, Professor, Department of Medicine, University of Sevilla, Avda. de Bellavista, 41014 Sevilla, Spain
S- Editor Gou SX L- Editor A E- Editor Xiong L