Background: One-third of patients with non-alcoholic fatty liver disease (NAFLD) develop dysmetabolic iron overload syndrome (DIOS), the pathogenesis of which is unknown. Altered production of the iron-regulatory peptide hepcidin has been reported in NAFLD, but it is unclear if this is related to iron accumulation, lipid status or steatohepatitis. Methods: Eighty-four patients with liver disease, 54 of which had iron overload, underwent liver biopsy (n = 66) and/or magnetic resonance imaging (n = 35) for liver iron content determination. Thirty-eight of the patients had NAFLD, 29 had chronic liver disease other than NAFLD, and 17 had untreated genetic hemochromatosis. Serum hepcidin was measured with ELISA in all patients and in 34 controls. Hepcidin antimicrobial peptide (HAMP) mRNA in liver tissue was determined with real-time-quantitative PCR in 36 patients. Results: Serum hepcidin was increased similarly in NAFLD with DIOS as in the other chronic liver diseases with iron overload, except for genetic hemochromatosis. HAMP mRNA in liver tissue, and serum hepcidin, both correlated to liver 2 2 iron content in NAFLD patients (r =0.45, p <0.05 and r =0.27, p < 0.05 respectively) but not to body mass index, NAFLD activity score or serum lipids. There was a good correlation between HAMP mRNA in liver tissue and serum hepcidin (r =0.39, p <0.01). Conclusions: In NAFLD with or without dysmetabolic iron overload, serum hepcidin and HAMP mRNA in liver correlate to body iron content but not to the degree of steatohepatitis or lipid status. Thus, the dysmetabolic iron overload syndrome seen in NAFLD is not caused by an altered hepcidin synthesis. Keywords: Hepcidin, Iron overload, Non-alcoholic fatty liver disease Background fibrosis, which are features of liver damage seen in Non-alcoholic fatty liver disease (NAFLD) is the most non-alcoholic steatohepatitis (NASH) which is the more prevalent liver disease worldwide, with an association to severe form of NAFLD [6–8]. obesity, insulin resistance and the metabolic syndrome The body’s iron balance is regulated by hepcidin, a 25 [1–3]. Approximately one-third of patients with NAFLD amino-acid peptide that inhibits iron uptake in the gut develop elevated serum ferritin and hepatic iron overload, and iron recycling from macrophages, consequently de- a condition known as the “dysmetabolic iron overload creasing iron levels in plasma . An inappropriately syndrome” (DIOS) [4, 5]. The underlying mechanisms for low hepcidin synthesis has been reported in NAFLD [10, DIOS are unknown. Increased iron stores could be of 11] which could facilitate iron uptake and predispose for pathogenic importance in NAFLD, since it may increase DIOS, but results are not consistent [12, 13]. Hepcidin the risk of hepatocyte ballooning, inflammation and levels in NAFLD are difficult to elucidate, since both obesity and diabetes may increase hepcidin production [12, 14, 15]. For example, in morbidly obese subjects, * Correspondence: email@example.com hepcidin is released from adipose tissue [12, 13, 15, 16], Unit of Liver Diseases, Department of Upper GI, C1-77 Huddinge, Karolinska University Hospital, Karolinska Institutet, 141 86 Stockholm, Sweden which may lead to anemia and entrapment of iron in Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Marmur et al. BMC Gastroenterology (2018) 18:78 Page 2 of 10 reticuloendothelial cells . Thus, in NAFLD data is together as CLD. NAFLD was defined as either Grade 1 or conflicting whether or not hepcidin predominantly cor- more steatosis in the liver biopsy according to Kleiner et al. relates to body iron stores [16, 17], to features of the , or a bright liver with increased echogenicity at ab- metabolic syndrome [18, 19] or the hepatic inflamma- dominal ultrasound investigation. Among the 17 patients tion seen in steatohepatitis (NASH). In a recent study, with HH, 12 were HFE C282Y homozygotes and five were hepatic iron measured by magnetic resonance imaging C282Y/H63D compound heterozygotes. In the group of 29 was found to be the major determinant of serum ferritin patients with chronic liver disease, eight had a normal iron in NAFLD . In a large study on individuals with content in the liver, and 21 had iron overload, and were metabolic syndrome, results suggested that that the iron classified as “chronic liver disease with iron overload” regulatory feedback on hepcidin synthesis was preserved (CLD-IO). One of these had received oral iron substitution in these patients . for several years; however, none had been treated with par- The aim of the present study was to elucidate whether enteral iron substitution or blood transfusions. Amongst body iron stores, steatohepatitis or lipid status in the 21 patients classified as CLD-IO, ten had a clinical NAFLD correlated to hepcidin synthesis. For this pur- phenotype of hemochromatosis (elevated serum ferritin pose, we compared serum hepcidin levels and hepatic and transferrin saturation, and hepatic iron overload) but hepcidin antimicrobial peptide (HAMP) gene expres- without homozygosity for the HFE C282Y mutation or sions in NAFLD patients with various degrees of iron compound heterozygosity for the C282Y and H63D muta- overload, to those of patients with other forms of ac- tions, and without alcohol overconsumption. The other 19 quired or genetic iron overload. We aimed to include CLD patients had alcohol overconsumption (> 30 g/day) patients with various hepcidin levels, and therefore we (n = 9), primary biliary cholangitis (n = 2), hepatitis C (n = included untreated hereditary hemochromatosis patients 1), alpha-1-antitrypsin deficiency (n = 1), porphyria cutanea (with a known hepcidin deficiency) as well as patients tarda (n = 1), cryptogenic cirrhosis (n =2), or with iron overload associated to other chronic liver dis- methotrexate-treated psoriasis (n = 3). None of the patients eases, presumably having elevated serum hepcidin levels. with HH, NAFLD or CLD with the clinical phenotype of We correlated our findings to iron indices, liver biopsy hemochromatosis had reported a previous or current alco- features, anthropometric data, and lipid parameters. hol consumption exceeding 20 g/day. Two CLD-IO pa- tients (with alpha-1-antitrypsin deficiency and alcohol Methods overconsumption, respectively) were heterozygous for the Patient data collection and investigations H63D mutation, and one (with alcohol overconsumption) All patients referred to the Unit of Liver Diseases at the was heterozygous for the C282Y mutation. Karolinska University Hospital for liver biopsy due to Iron overload was defined as a histologic iron score of chronic liver disease and/or hemochromatosis, and with ≥1 or an estimated iron content > 40 μmol/g on mag- an elevated serum ferritin, between January 2008 and netic resonance imaging (MRI) investigation (see below). April 2013, were asked to participate in the study. The patient groups are displayed in Table 1. Hyperferritinemia was defined as a serum ferritin > Liver biopsy was performed in 66 out of the 84 pa- 350 μg/L. In addition, patients with chronic liver disease tients. MRI was used for iron assessment in 35 cases, and normal iron parameters undergoing liver biopsy and in 14 of these, histology was lacking. In 21 cases, were enrolled for comparison. All patients were over there was both histology and MRI. In four cases (two 18 years of age and had given written informed consent. HH homozygotes, one HH compound heterozygote, and One patient was excluded due to iron deficiency. No pa- one with CLD and normal ferritin) both liver histology tients included had been subject to treatment with iron and MRI was lacking. reduction therapy before entering the study. A total of 84 patients were enrolled (26 females, 58 Data collection from controls males), of which 62 had elevated ferritin levels and 23 a A total of 40 healthy controls, recruited from hospital normal serum ferritin concentration. Thirty-eight of the 84 staff, participated in the study. None had a history of patients had NAFLD, 17 had untreated hereditary liver disease. Written consent was given. Of the controls, hemochromatosis (HH), and 29 had various other causes six individuals were excluded (elevated transaminases in of chronic liver disease (CLD), such as autoimmune liver one case, compound heterozygosity of the HFE gene and disease, alcoholic liver disease, chronic viral hepatitis, elevated serum ferritin in one case, iron deficiency with alpha-1-antitrypsin deficiency, cryptogenic cirrhosis, por- serum ferritin < 15 μg/L in three cases, and elevated fer- phyria cutanea tarda, methotrexate-induced liver fibrosis, ritin (413 μg/L) in one case). The remaining 34 controls or the hemochromatosis phenotype but without the C282Y were included in the study (Table 1). or H63D mutations. All these other etiologies (except Biochemical data was collected at the time of enroll- NAFLD and HFE-associated HH) were thus grouped ment in the study. Blood samples were drawn before 10 Marmur et al. BMC Gastroenterology (2018) 18:78 Page 3 of 10 Table 1 Clinical and laboratory data for patients and controls Control NAFLD NAFLD with DIOS CLD CLD-IO Compound heterozygous HH Homozygous HH N =34 N =22 N =16 N =8 N =21 N =5 N =12 Gender (F/M) 19/15 8/14 5/11 4/4 6/15 1/4 2/10 Age (y) 40 ± 10* 54 ± 16 59 ± 10 57 ± 8 58 ± 8 59 ± 9 51 ± 6 BMI (kg/m ) 23.3 ± 2.6* 30.4 ± 4.2# 28.1 ± 2.4 27.3 ± 5.1 27.0 ± 4.2 29.0 ± 3.5 28.2 ± 4.8 Hemoglobin (g/L) 142 ± 11 150 ± 15 149 ± 17 140 ± 10 140 ± 17 158 ± 15 154 ± 12 ALT (U/L) 18 ± 6* 94 ± 76 59 ± 41 71 ± 59 53 ± 29 41 ± 35 82 ± 35 CRP (mg/L) 1.1 ± 0.4 3.0 ± 2.9 1.8 ± 0.9 2.8 ± 3.0 3.8 ± 5.5 3.6 ± 4.0 3.3 ± 2.7 Serum ferritin (μg/L) 94 ± 87 304 ± 248 816 ± 285¤ 454 ± 688 1304 ± 1295¤ 878 ± 408¤ 1753 ± 998¤ Transferrin saturation (%) 0.28 ± 0.11 0.28 ± 0.09 0.39 ± 0.09 0.31 ± 0.15 0.50 ± 0.21§ 0.46 ± 0.10§ 0.76 ± 0.21* Hepatic iron score N.D. 0.11 ± 0.21 2.19 ± 0.95¤ 0.29 ± 0.27 2.98 ± 1.22¤ 3.38 ± 0.75¤ 4.35 ± 0.63¤ CLD chronic liver disease, IO iron overload, HH hereditary hemochromatosis, NAFLD non-alcoholic fatty liver disease, DIOS dysmetabolic iron overload syndrome. Values denote mean ± S.D *P < 0.05 vs. all other groups #p < 0.05 vs. Control, CLD, CLD-IO ¤p < 0.05 vs. Control, NAFLD, CLD §p < 0.05 vs. Control, NAFLD, DIOS, CLD, homozygous HH A.M. in the morning. Subjects were not fasting but had Briefly, plasma/serum was diluted 1:4 using Bio-plex sam- had a light breakfast. Routine blood chemistry analyses ple diluents. To obtain the nine point (including blank) as well as HFE mutation analysis were performed on all standard curve, the kit standard was reconstituted and di- subjects at the Department of Clinical Chemistry at Kar- luted fourfold. The 10× IL-6 and TNFα coupled beads olinska University Hospital. Body mass index was calcu- were diluted in kit assay buffer and added to all standard lated using the formula: weight in kilogram / (height in and sample wells. The plate was incubated on shaker meters) . 30 min. After washing IL-6 and TNFα biotinylated detec- tion antibodies were added and the plate was incubated as Quantitative assay of hepcidin in serum samples above. In the final step, PE-conjugated Streptavidin was Freshly drawn samples from the 84 patients and 34 con- added and the plate was run on a Magpix instrument trols were centrifuged and serum was stored at − 70 °C (Luminex Corporation, Austin TX, USA) and analyzed until analysis. Samples were analyzed for hepcidin by a with xPonent software (Luminex). competitive ELISA kit (Bachem, Peninsula Laboratories, LLC, CA, United States) as reported previously . Analysis of HAMP mRNA in liver biopsies Reference ranges established in 83 normal subjects Sixty-six patients underwent liver biopsy, and tissue showed hepcidin levels that ranged 8–76 and 2–50 μg/L from 39 of these was collected for hepcidin mRNA ana- for men and women, respectively (2.5–97.5 percentiles). lysis. Tissues were immediately immersed in RNAlater The results were significantly different between genders. and stored at − 70°C until processed. Total RNA was As internal controls, pooled sera of 7 and 6 samples successfully retrieved from 36 of the 39 utilized liver bi- representing low (≈0.4 μg/L) and normal (≈3 μg/L) levels opsies with a dry weight of 0.3–5.9 mg using the RNA- respectively were frozen at − 70 °C. Control sera were queous -4PCR kit (Ambion PN AM1914). Recovered run in 6 replicates at each assay. The intra-assay vari- quantities of RNA ranged from 13 to 200 ng/μL. The ation showed CVs of 18% for low and 13% for normal quality and quantity of the extracted RNA was verified controls, while inter-assay CVs were 18 and 19%, re- with the Bio-Rad Experion 700–7000 electrophoresis spectively. The lower limit of detection, calculated as 3 system, and only samples with an RQI > 8 were included SD above the lowest standard, was 0.05 μg/L and linear- in the study. cDNA synthesis was carried out with the ity for this kit was determined as between 0,2–5 μg/L High Capacity Reverse Transcriptase Kit (Applied Bio- (2–50 μg/L for samples diluted 1:10). Samples outside systems), using 65–930 ng of total RNA per sample. De- linearity limits were rerun using proper dilution factor, termination of specific mRNA levels was performed as and all samples were run in duplicate. described previously . Analysis of IL-6 and TNFα Histologic examination of liver biopsy samples IL-6 and TNFα were measured using Bio-plex Pro Human Liver biopsy samples were revalued by an experienced Cytokine Group 1 kit (Bio-rad Laboratories, Hercules, pathologist (O.D.) blinded to clinical data. Samples from CA, USA) according to the manufacturer’s instructions. NAFLD-patients were evaluated for the degree of Marmur et al. BMC Gastroenterology (2018) 18:78 Page 4 of 10 steatosis (0–3), lobular inflammation (0–3) and hepato- in 21 of these (Fig. 1). The liver iron was assessed cellular ballooning (0–2) according to Kleiner et al. . semi-quantitatively as described by Gandon et al. . The unweighted sum of these three variables were used In the correlation analyses of serum hepcidin to liver iron to calculate the NAFLD activity score (NAS). Patients content, MRI iron was approximated to histologic liver with NAS ≥5 were diagnosed with NASH. iron (HIS) score based on the correlation estimated from Siderosis was determined for all patients Fig. 1:<40 μmol iron/g tissue = HIS 0; 40–74 μmol/g = semi-quantitatively on histopathologic examination of HIS 1; 75–129 μmol/g = HIS 2; 130–179 μmol/g = HIS 3; Perls’ stained liver biopsy samples adapted from Deugnier 180–239 μmol/g = HIS 4; ≥240 μmol iron/g tissue = HIS 5. et al.  to match available levels of magnification. An iron score from 0 to 4 for iron in hepatocytes was Statistical analyses determined as follows:  granules absent or barely dis- The relationship between two categorical variables was cernible at a magnification of 400X;  barely discern- examined with Chi -test or Fisher’s exact test (when ap- ible granules at a magnification of 200X but easily plicable). Numerical values of laboratory parameters confirmed at a magnification of 400X;  discrete gran- were analyzed using one-way ANOVA and validated for ules at 100X magnification;  discrete granules easily equal variance and normal distribution. Kruskal-Wallis confirmed at magnification of 40X, but barely discernible ANOVA was used when the assumptions did not hold. at a magnification of 20X;  granules obvious at a mag- The correlation between two numerical variables was nification of 20X, and barely visible for the naked eye. analyzed with simple linear regression validated for lin- RES-iron was also determined and scored as  none, earity, variance between observations and for normal  mild,  or more than mild, as described by Nelson distribution. In the cases where the assumptions did not et al. . These two scores were transformed into a hold the Spearman’s rank order correlation was used in- histologic iron score (HIS) ranging from 0 to 5, compris- stead. Multiple linear regression was used for variables ing the score for iron in hepatocytes (0–4), plus one that were significantly correlated to serum hepcidin in point for RES iron in those cases where it had been de- the simple linear regression. A p-value < 0.05 was con- termined as more than mild, or a half point where it has sidered statistically significant. been determined as mild. Iron overload was defined as a histologic iron score of ≥1. Results Clinical and laboratory data The distribution of patients, and clinical and laboratory Magnetic resonance imaging (MRI) data of patients and controls are demonstrated in MRI was used for detection and quantification of liver Table 1. Controls were significantly younger than pa- iron overload in 35 patients and correlated to histology tients, and had lower BMI, ALT and serum ferritin Fig. 1 Graph demonstrating the correlation between MRI iron content (μmol/g) and histological iron score in 21 patients in whom both MRI and liver biopsy was performed. There was a good correlation between these variables (r = 0.77; p < 0.01) Marmur et al. BMC Gastroenterology (2018) 18:78 Page 5 of 10 levels. BMI was highest in the NAFLD patient group. homozygous HH. The hepcidin/iron score ratio was Transferrin saturation was significantly increased in pa- slightly lower in those with ALD or hepatitis C (18.7 tients with homozygous HH, and in the 10 CLD-IO pa- ±8.1) as compared with those without alcohol overcon- tients with a HH phenotype without HFE mutations, sumption (22.4±10.2), or DIOS (30.8±23.7), however not compared with the other patient groups and controls. statistically significant. There was a significant correl- Hepatic iron score did not differ significantly between ation between serum hepcidin levels and hepatic HAMP patients with DIOS and CLD-IO. mRNA (r = 0.39, p < 0.01). Distribution of HFE mutations Clinical, laboratory and histological findings in patients The distribution of HFE mutations are shown in Table 2. with NAFLD with or without DIOS (Table 3) Among patients with chronic liver disease and iron over- Serum hepcidin, serum transferrin saturation and hepatic load (CLD-IO), four were heterozygous for C282Y, two iron score were all significantly higher in NAFLD with homozygous and one heterozygous for H63D. The H63D DIOS as compared with NAFLD without DIOS (p <0.05). mutation was significantly more frequent in patients with Serum levels of triglycerides or total cholesterol did not NAFLD as compared with the controls (p <0.05). differ significantly between the groups. Levels of TNF-α and IL-6 were highest in NAFLD without DIOS and ele- Correlation analysis of histologic iron score and hepatic vated serum ferritin (difference not statistically signifi- iron content determined by MRI cant). HAMP mRNA in liver tissue correlated to the Simple linear regression showed a good correlation hepatic iron score (r =0.45, p < 0.05) but not to NAFLD between histologic iron score and hepatic iron con- activity score (r =0.003, p < 0.89). Serum hepcidin corre- tent determined by MRI, as demonstrated in Fig. 1 lated significantly to serum ferritin (r =0.20, p < 0.01) and 2 2 (r =0.77, p < 0.01). serum transferrin saturation (r =0.17, p <0.01) but not to BMI, TNF-α, IL-6, triglycerides or cholesterol. In multiple Serum hepcidin and hepcidin mRNA in liver biopsies linear regression analysis only ferritin correlated signifi- Serum hepcidin values for the different patient groups and cantly to serum hepcidin levels when adjusted for other controls are shown in Fig. 2. Serum hepcidin levels were variables. There was no significant difference in stage of fi- significantly increased in patients with NAFLD with DIOS brosis, grade of steatosis, ballooning, lobular inflammation and in those with chronic liver disease with iron overload or NAFLD activity score between the groups (Table 3). (CLD-IO) compared with the other groups. The ratios be- tween serum hepcidin and ferritin are shown in Fig. 3.As Discussion expected, this ratio was significantly reduced in homozy- In the present study, we demonstrate that in NAFLD pa- gous HH compared with the other groups. Among pa- tients, hepcidin in serum and HAMP mRNA in liver tissue tients with CLD-IO, this ratio was slightly lower in those correlate significantly to body iron stores, regardless if with alcoholic liver disease (ALD) or hepatitis C (0.049 they are expressed as serum ferritin or liver iron content. ±0.034) as compared with those without alcohol overcon- Furthermore, there was no correlation to the degree of sumption (0.058±0.032), or DIOS patients (0.070±0.037), steatohepatitis (defined as NAFLD activity score), to lipid although these differences were not statistically significant. parameters (serum cholesterol or triglycerides), body mass Figure 4 shows the ratios between serum hepcidin and index, or C-reactive protein. We found that serum hepci- hepatic iron score, which was similar in patients with din levels in NAFLD patients with dysmetabolic iron over- CLD-IO and DIOS, and reduced in those with load (DIOS) are similar to those found in other liver Table 2 HFE genotypes in patients and controls wt/wt C282Y/wt C282Y/C282Y C282Y/H63D H63D/wt H63D/H63D Controls (n = 34) 26 4 –– 4 – NAFLD with normal iron stores (n = 22) 13 1 –– 6* 1 NAFLD with DIOS (n = 16) 12 2 –– 2 – CLD (n =8) 4 –– – 2 – CLD-IO (n = 21) 14 4 –– 12 Compound heterozygous HH (n =5) –– – 5 –– Homozygous HH (n = 12) –– 12 –– – CLD chronic liver disease, IO iron overload, HH hereditary hemochromatosis, NAFLD non-alcoholic fatty liver disease, DIOS dysmetabolic iron overload *p < 0.05 in patients with NAFLD vs. controls one missing value two missing values Marmur et al. BMC Gastroenterology (2018) 18:78 Page 6 of 10 Fig. 2 Serum hepcidin levels (μg/L) in the different patient groups. The box plots show the median, the interquartile range and the min-max values. Hepcidin levels were significantly increased in chronic liver disease with iron overload (CLD-IO) and non-alcoholic fatty liver disease with dysmetabolic iron overload (NAFLD-DIOS) compared with the other groups (Kruskal-Wallis ANOVA, p < 0.05) diseases with iron overload (CLD-IO), except for her- alcoholic liver disease and hepatitis C had a trend to editary hemochromatosis, in which patients have an somewhat lower levels. Others have demonstrated that inherited hepcidin deficiency. In our patient cohort hepatic iron is the major determinant of serum ferritin without morbid obesity, hepatic HAMP mRNA levels levels in NAFLD, results in line with the present study showed a good correlation to the serum hepcidin values . Together, these findings point at an adequate hep- measured by ELISA. When calculating the hepcidin cidin synthesis in NAFLD in relation to iron stores, and levels in relation to serum ferritin (Fig. 3) or to the liver the iron accumulation in DIOS cannot be explained by iron score (Fig. 4), patients with DIOS had overall simi- hepcidin deficiency, in contrast to what is seen in lar ratios as patients with CLD-IO, although those with hereditary hemochromatosis. Fig. 3 The ratios between serum hepcidin (μg/L) and serum ferritin (μg/L). Patients with homozygous HH had significantly lower ratios compared with the other groups (Kruskal-Wallis ANOVA, p < 0.05) Marmur et al. BMC Gastroenterology (2018) 18:78 Page 7 of 10 Fig. 4 The ratios between serum hepcidin (μg/L) and hepatic iron contents (“iron score”). The calculation of iron scores is described in Methods. Patients with homozygous HH had significantly lower ratios compared with the other groups (Kruskal-Wallis ANOVA, p < 0.05) Some other previous studies have presented conflicting 21, 27, 28]. Senates et al. found an association between results. Barisani et al. found an inadequate hepcidin pro- serum hepcidin and cholesterol and triglycerides levels, duction for a given level of iron status in NAFLD pa- but not with iron parameters , which contrasts with tients compared to controls, although not as low as in our findings. In obesity, hepcidin can be produced by beta-thalassemia or hereditary hemochromatosis . In adipose tissue [13, 28], possibly through activation of contrast, several other studies found hepcidin levels to hemojuvelin gene expression . Thus, in morbidly correlate to iron parameters in NAFLD and DIOS [17, obese patients undergoing bariatric surgery, hepcidin Table 3 Clinical, laboratory and liver biopsy findings in patients with NAFLD with and without dysmetabolic iron overload syndrome (DIOS), and normal vs. elevated serum ferritin, respectively NAFLD without DIOS (n = 22) NAFLD with DIOS With serum ferritin < 350 μg/L (n = 15) With serum ferritin > 350 μg/L (n =7) (n = 16) Serum hepcidin (μg/L) 24 ± 19 37 ± 13 53 ± 28* Serum ferritin (μg/L) 156 ± 78* 621 ± 170 816 ± 285 Transferrin saturation (%) 28 ± 8 25 ± 10 39 ± 9# Liver iron score 0.03 ± 0.13 0.14 ± 0.24 2.13 ± 0.92* CRP (mg/L) 2.7 ± 2.2 3.7 ± 4.2 1.8 ± 0.91 Plasma-triglycerides (mmol/L) 2.89 ± 1.09 1.83 ± 1.09 1.95 ± 0.90 Plasma cholesterol (mmol/L) 5.18 ± 0.96 5.25 ± 0.84 5.25 ± 0.71 TNF-α (ng/L) 6.27 ± 5.13 137 ± 316 8.37 ± 5.82 IL-6 (ng/L) 2.47 ± 1.71 32.3 ± 62.7 2.36 ± 1.02 NAS 4.5 ± 1.8 3.6 ± 1.7 4.4 ± 1.8 Steatosis (grade) 2.07 ± 0.80 2.00 ± 1.00 2.72 ± 0.65 Ballooning 1.07 ± 0.59 0.80 ± 0.84 0.63 ± 0.67 Lobular inflammation 1.20 ± 0.86 1.00 ± 0.71 1.00 ± 1.00 Portal inflammation 0.27 ± 0.46 0.60 ± 0.55 0.18 ± 0.40 Fibrosis 1.33 ± 0.90 2.00 ± 1.00 2.72 ± 0.65 Values denote mean ± S.D *=p < 0.05 (vs. the other groups) #= p < 0.05 (vs. NAFLD with serum ferritin > 350 μg/L) Marmur et al. BMC Gastroenterology (2018) 18:78 Page 8 of 10 levels correlate to the grade of obesity, but not to the de- group in the present study, relative to the patient cohort, gree of fat in the liver tissue [12, 15]. Likewise, the pres- is a limitation when comparing serum hepcidin levels in ence of NASH did not alter the expression of HAMP liver disease patients and healthy controls. mRNA in adipose tissue . Low-grade inflammation Future studies need to focus on hepcidin-independent associated with obesity could lead to elevation of both mechanisms for the iron-loading seen in NAFLD with serum ferritin and hepcidin levels. In inflammatory con- DIOS. Hitherto published data indicate that activated ditions, elevated serum hepcidin would diminish iron iron regulatory protein-1 and increased expression of uptake and mobilization, possibly causing entrapment of duodenal divalent metal transporter-1 have been found iron in Kupffer cells . However, none of our patients in NASH . Also, bone morphogenic protein-binding were morbidly obese, and the strong correlation between endothelial regulator  and hepatocyte nuclear hepcidin in serum and HAMP mRNA in liver tissue in factor-4alpha  have been reported to influence iron the present study indicates a negligible contribution absorption, and in mice, a high fat diet by itself could in- from adipose tissue to hepcidin synthesis in our cohort. crease iron absorption . An impairment in the ability It has been reported that hepcidin levels were depressed of hepcidin to inhibit iron absorption was demonstrated in patients with chronic hepatitis C  or alcoholic liver in DIOS, suggesting hepcidin resistance in this condition disease , suggesting that hepcidin deficiency play a role . Nevertheless, it is unknown if the iron loading seen in the iron accumulation seen in these conditions. As com- in up to one-third of patients with NAFLD is a conse- pared to NAFLD-DIOS in our cohort, we found a some- quence of the altered lipid metabolism, or an altered ex- what lower hepcidin-to-ferritin and hepcidin-to-iron score pression of iron-regulatory genes, or a combination of ratios in patients with alcoholic liver disease and hepatitis both. This topic warrants future research. C, although the present study was underpowered to detect atruedifferenceinthisregard. This findingisinline with Conclusions the view that there is an adequate hepcidin synthesis in In conclusion, we found that in patients with non-alcoholic NAFLD-DIOS, why other explanations for the iron accu- fatty liver disease with or without dysmetabolic iron over- mulation in this condition have to be sought for . load, serum hepcidin correlates to iron indices such as We did not find an increased frequency of C282Y or serum ferritin, transferrin saturation and liver iron con- H63D mutations in NAFLD patients with DIOS as com- tents, but not to body mass index, NAFLD activity score, or pared to patients with other liver diseases, or healthy lipid parameters. Hepcidin levels in NAFLD-DIOS are simi- controls. However, the H63D mutation was enriched in lar to those found in other liver diseases with iron overload, NAFLD patients with normal iron stores, indicating that except for genetic hemochromatosis. These data indicate this mutation may play a role in NAFLD pathogenesis, that NAFLD-DIOS is a condition with an adequate hepci- as suggested previously . din synthesis and preserved iron-regulatory feedback. Eighteen of our 84 patients did not agree to undergo Abbreviations liver biopsy. In these cases, iron assessment was instead ALD: Alcoholic liver disease; CLD: Chronic liver disease; CLD-IO: Chronic liver performed by magnetic resonance imaging (MRI), which disease with iron overload; DIOS: Dysmetabolic iron overload syndrome; HAMP: Hepcidin antimicrobial peptide; HH: Hereditary hemochromatosis; is considered to be an accurate method to quantify iron MRI: Magnetic resonance imaging; NAFLD: Non-alcoholic fatty liver disease; overload in the range 60–375 μmol/g . It is not in- NASH: Non-alcoholic steatohepatitis fluenced by steatosis or fibrosis and in patients with cir- Acknowledgements rhosis it may be even more accurate than biopsy . In We are grateful to Terri Lindholm for MRI iron quantification expertise, and 21 cases, we performed both liver biopsy and MRI, Pia Loqvist, Ingrid Ackzell, and Eva Berglund for blood and tissue sampling obtaining a good correlation in cases with a hepatic iron and excellent patient care. score of 2 or more. Funding The strength of the present study is that hepcidin was This study was supported by grants from the Swedish Society of Medicine measured both in serum and as HAMP mRNA in liver (Bengt Ihre’s fund and Swedish Gastroenterology Fund), the Karolinska Institutet (Ruth and Richard Julins Foundation) and from the Stockholm tissue, the correlations of which were excellent. Further- County Council (ALF project 20150403). more, iron content was assessed both with MRI and liver biopsy, and NAFLD patients were compared with other Availability of data and materials The datasets generated in the current study are available from the patient cohorts with various degree of iron overload, in- corresponding author on reasonable request. cluding genetic hemochromatosis who has an inherited hepcidin deficiency. The major limitation of the study is Authors’ contributions the small cohort, making it underpowered to perform Study conception and design: PS, JM. Acquisition of data: JM, PS, SB, GE, LO, NA, OD. Statistical analysis: PS. Analysis and interpretation of data: All authors. sub-analyses of various patient groups, e.g. NAFLD Drafting of manuscript: JM, PS. Critical revision: All authors. Guarantor of without DIOS, alcoholic liver disease or hepatitis C. article: PS. All authors approved the final version of the article, including the Also, the smaller size and the younger age of the control authorship list. Marmur et al. BMC Gastroenterology (2018) 18:78 Page 9 of 10 Ethics approval and consent to participate 14. Jiang F, Sun ZZ, Tang YT, Xu C, Jiao XY. Hepcidin expression and iron The study was conducted in accordance with the Helsinki Declaration of parameters change in type 2 diabetic patients. Diabetes Res Clin Pract. 1975, as revised in 1983, and approved by the Ethics Committee at 2011;93(1):43–8. Karolinska University Hospital, Stockholm, Sweden (No. 2007/1297–31/2). All 15. Auguet T, Aragones G, Berlanga A, Martinez S, Sabench F, Binetti J, Aguilar patients signed the informed consent to participate in the study. C, Porras JA, Molina A, Del Castillo D, et al. Hepcidin in morbidly obese women with non-alcoholic fatty liver disease. PLoS One. 2017;12(10): e0187065. Competing interests 16. Ruivard M, Laine F, Ganz T, Olbina G, Westerman M, Nemeth E, Rambeau M, The authors declare that they have no competing interests. Mazur A, Gerbaud L, Tournilhac V, et al. Iron absorption in dysmetabolic iron overload syndrome is decreased and correlates with increased plasma hepcidin. J Hepatol. 2009;50(6):1219–25. Publisher’sNote 17. Nelson JE, Brunt EM, Kowdley KV. Nonalcoholic steatohepatitis clinical Springer Nature remains neutral with regard to jurisdictional claims in research N: lower serum hepcidin and greater parenchymal iron in published maps and institutional affiliations. nonalcoholic fatty liver disease patients with C282Y HFE mutations. Hepatology. 2012;56(5):1730–40. Author details 18. SenatesE,Yilmaz Y, ColakY,OzturkO, Altunoz ME,Kurt R,OzkaraS, Unit of Liver Diseases, Department of Upper GI, C1-77 Huddinge, Karolinska Aksaray S, Tuncer I, Ovunc AO. Serum levels of hepcidin in patients University Hospital, Karolinska Institutet, 141 86 Stockholm, Sweden. Unit of with biopsy-proven nonalcoholic fatty liver disease. Metab Syndr Relat Gastroenterology and Hepatology, Department of Medicine, Ersta Hospital, Disord. 2011;9(4):287–90. Karolinska Institutet, Stockholm, Sweden. Unit of Clinical Chemistry, 19. Aigner E, Weiss G, Datz C. Dysregulation of iron and copper homeostasis in Department of Laboratory Medicine, Karolinska University Hospital, Karolinska nonalcoholic fatty liver. World J Hepatol. 2015;7(2):177–88. Institutet, Stockholm, Sweden. Department of Radiology, Ersta Hospital, 20. Ryan JD, Armitage AE, Cobbold JF, Banerjee R, Borsani O, Dongiovanni Karolinska Institutet, Stockholm, Sweden. Unit of Pathology, Department of P, Neubauer S, Morovat R, Wang LM, Pasricha SR, et al. Hepatic iron is Laboratory Medicine, Karolinska University Hospital, Karolinska Institutet, the major determinant of serum ferritin in NAFLD patients. Liver Int. Stockholm, Sweden. 2018;38(1):164–73. 21. Martinelli N, Traglia M, Campostrini N, Biino G, Corbella M, Sala C, Busti F, Received: 13 March 2018 Accepted: 23 May 2018 Masciullo C, Manna D, Previtali S, et al. Increased serum hepcidin levels in subjects with the metabolic syndrome: a population study. PLoS One. 2012; 7(10):e48250. 22. Kleiner DE, Brunt EM. Nonalcoholic fatty liver disease: pathologic patterns References and biopsy evaluation in clinical research. Semin Liver Dis. 2012;32(1):3–13. 1. Vernon G, Baranova A, Younossi ZM. Systematic review: the epidemiology 23. Dahlfors G, Stal P, Hansson EC, Barany P, Sisowath C, Onelov L, Nelson D, and natural history of non-alcoholic fatty liver disease and non-alcoholic Eggertsen G, Marmur J, Beshara S. Validation of a competitive ELISA assay steatohepatitis in adults. Aliment Pharmacol Ther. 2011;34(3):274–85. for the quantification of human serum hepcidin. Scand J Clin Lab Invest. 2. Marchesini G, Brizi M, Bianchi G, Tomassetti S, Bugianesi E, Lenzi M, 2015;75(8):652–8. McCullough AJ, Natale S, Forlani G, Melchionda N. Nonalcoholic fatty liver 24. Deugnier Y, Turlin B. Pathology of hepatic iron overload. World J disease: a feature of the metabolic syndrome. Diabetes. 2001;50(8):1844–50. Gastroenterol. 2007;13(35):4755–60. 3. Loomba R, Sanyal AJ. The global NAFLD epidemic. Nat Rev Gastroenterol 25. Nelson JE, Wilson L, Brunt EM, Yeh MM, Kleiner DE, Unalp-Arida A, Kowdley Hepatol. 2013;10(11):686–90. KV. Nonalcoholic steatohepatitis clinical research N: relationship between 4. Nelson JE, Klintworth H, Kowdley KV. Iron metabolism in nonalcoholic fatty the pattern of hepatic iron deposition and histological severity in liver disease. Curr Gastroenterol Rep. 2012;14(1):8–16. nonalcoholic fatty liver disease. Hepatology. 2011;53(2):448–57. 5. Datz C, Felder TK, Niederseer D, Aigner E. Iron homeostasis in the metabolic 26. Gandon Y, Olivie D, Guyader D, Aube C, Oberti F, Sebille V, Deugnier Y. syndrome. Eur J Clin Investig. 2013;43(2):215–24. Non-invasive assessment of hepatic iron stores by MRI. Lancet. 2004; 6. Manousou P, Kalambokis G, Grillo F, Watkins J, Xirouchakis E, Pleguezuelo M, 363(9406):357–62. Leandro G, Arvaniti V, Germani G, Patch D, et al. Serum ferritin is a 27. Handa P, Vemulakonda AL, Maliken BD, Morgan-Stevenson V, Nelson JE, discriminant marker for both fibrosis and inflammation in histologically Dhillon BK, Hennessey KA, Gupta R, Yeh MM, Kowdley KV. Differences in proven non-alcoholic fatty liver disease patients. Liver Int. 2011;31(5):730–9. hepatic expression of iron, inflammation and stress-related genes in 7. Kowdley KV, Belt P, Wilson LA, Yeh MM, Neuschwander-Tetri BA, Chalasani patients with nonalcoholic steatohepatitis. Ann Hepatol. 2017;16(1):77–85. N, Sanyal AJ, Nelson JE, Network NCR. Serum ferritin is an independent 28. Coimbra S, Catarino C, Santos-Silva A. The role of adipocytes in the predictor of histologic severity and advanced fibrosis in patients with modulation of iron metabolism in obesity. Obes Rev. 2013;14(10):771–9. nonalcoholic fatty liver disease. Hepatology. 2012;55(1):77–85. 29. Luciani N, Brasse-Lagnel C, Poli M, Anty R, Lesueur C, Cormont M, 8. Fracanzani AL, Valenti L, Bugianesi E, Vanni E, Grieco A, Miele L, Consonni D, Laquerriere A, Folope V, LeMarchand-Brustel Y, Gugenheim J, et al. Fatta E, Lombardi R, Marchesini G, et al. Risk of nonalcoholic steatohepatitis Hemojuvelin: a new link between obesity and iron homeostasis. Obesity. and fibrosis in patients with nonalcoholic fatty liver disease and low visceral 2011;19(8):1545–51. adiposity. J Hepatol. 2011;54(6):1244–9. 30. Fujita N, Sugimoto R, Takeo M, Urawa N, Mifuji R, Tanaka H, Kobayashi Y, 9. Corradini E, Pietrangelo A. Iron and steatohepatitis. J Gastroenterol Hepatol. Iwasa M, Watanabe S, Adachi Y, et al. Hepcidin expression in the liver: 2012;27(Suppl 2):42–6. relatively low level in patients with chronic hepatitis C. Mol Med. 2007;13(1– 10. Siddique A, Nelson JE, Aouizerat B, Yeh MM, Kowdley KV, Network NCR. Iron 2):97–104. deficiency in patients with nonalcoholic fatty liver disease is associated with 31. Harrison-Findik DD, Schafer D, Klein E, Timchenko NA, Kulaksiz H, Clemens obesity, female gender, and low serum hepcidin. Clin Gastroenterol D, Fein E, Andriopoulos B, Pantopoulos K, Gollan J. Alcohol metabolism- Hepatol. 2014;12(7):1170–8. mediated oxidative stress down-regulates hepcidin transcription and leads 11. Barisani D, Pelucchi S, Mariani R, Galimberti S, Trombini P, Fumagalli D, to increased duodenal iron transporter expression. J Biol Chem. 2006; Meneveri R, Nemeth E, Ganz T, Piperno A. Hepcidin and iron-related gene 281(32):22974–82. expression in subjects with Dysmetabolic hepatic Iron overload. J Hepatol. 32. Olthof AW, Sijens PE, Kreeftenberg HG, Kappert P, van der Jagt EJ, Oudkerk 2008;49(1):123–33. M. Non-invasive liver iron concentration measurement by MRI: comparison 12. Vuppalanchi R, Troutt JS, Konrad RJ, Ghabril M, Saxena R, Bell LN, Kowdley of two validated protocols. Eur J Radiol. 2009;71(1):116–21. KV, Chalasani N. Serum hepcidin levels are associated with obesity but not liver disease. Obesity. 2014;22(3):836–41. 33. Hoki T, Miyanishi K, Tanaka S, Takada K, Kawano Y, Sakurada A, Sato M, Kubo 13. Bekri S, Gual P, Anty R, Luciani N, Dahman M, Ramesh B, Iannelli A, Staccini- T, Sato T, Sato Y, et al. Increased duodenal iron absorption through up- Myx A, Casanova D, Ben Amor I, et al. Increased adipose tissue expression regulation of divalent metal transporter 1 from enhancement of iron of hepcidin in severe obesity is independent from diabetes and NASH. regulatory protein 1 activity in patients with nonalcoholic steatohepatitis. Gastroenterology. 2006;131(3):788–96. Hepatology. 2015;62(3):751–61. Marmur et al. BMC Gastroenterology (2018) 18:78 Page 10 of 10 34. Hasebe T, Tanaka H, Sawada K, Nakajima S, Ohtake T, Fujiya M, Kohgo Y. Bone morphogenetic protein-binding endothelial regulator of liver sinusoidal endothelial cells induces iron overload in a fatty liver mouse model. J Gastroenterol. 2017;52(3):341-351. 35. Shi W, Wang H, Zheng X, Jiang X, Xu Z, Shen H, Li M. HNF-4alpha negatively regulates Hepcidin expression through BMPR1A in HepG2 cells. Biol Trace Elem Res. 2017;176(2):294–304. 36. Dongiovanni P, Lanti C, Gatti S, Rametta R, Recalcati S, Maggioni M, Fracanzani AL, Riso P, Cairo G, Fargion S, et al. High fat diet subverts hepatocellular iron uptake determining dysmetabolic iron overload. PLoS One. 2015;10(2):e0116855. 37. Rametta R, Dongiovanni P, Pelusi S, Francione P, Iuculano F, Borroni V, Fatta E, Castagna A, Girelli D, Fargion S, et al. Hepcidin resistance in dysmetabolic iron overload. Liver Int. 2016;36(10):1540–8.
– Springer Journals
Published: Jun 5, 2018