Phenotypic Variability in Genetic Hemochromatosis: Study of Five SNPs in a Cohort of 342 Compound Heterozygotes for the HFE Cys282Tyr / His63Asp Mutations

The HFE p.Cys282Tyr (C282Y) mutation is the disease-causing defect in most cases of genetic hemochromatosis, but involvement of the p.His63Asp (H63D) variant in the pathogenesis of the illness is still a matter of debate. Some authors consider C282Y/H63D (YD) subjects as at risk of developing hemochromatosis, whereas others argue that iron overload, seen in some of these subjects, is related to external factors. As for the C282Y homozygotes, genetic modifiers might contribute to the phenotypic variability of hemochromatosis in YD subjects. This work was thus aimed at assessing the possible influence of Single Nucleotide Polymorphism (SNPs) suspected to play a role in iron metabolism on disease expression in compound heterozygotes. We genotyped five SNPs in 342 YD unrelated Caucasian patients (208 males, 134 females) from South France. Information on demographics, clinical manifestations, laboratory tests, viral status and alcohol consumption was obtained from the medical records of the participants. The selected SNPs were already described as modifiers of iron overload by previous studies. They are located in the TMPRSS6 (rs855791 and rs4820268), TF (rs3811647), CYBRD1 (rs884409) and GNPAT (rs11558492) genes. The role of GNPAT in iron metabolism is unknown but it has been recently associated with the severity of the clinical expression in C282Y homozygous subjects1. Analysis of the SNPs was performed using Endpoint TaqMan Genotyping technology. For each SNP, the additive effect of the minor allele on the logarithm of serum ferritin (SF) was first assessed through a linear regression, controlling for age, sex, alcohol consumption, and patient's referral motive. We found a marginally significant association between rs884409 allele G of CYBRD1 and SF (p = 0.014). In a second step, we restricted our analysis to the 77 most severe males (SF > 500 µg/l and no past or current history of alcohol consumption > 20g/day), and found that the GNPAT rs11558492 allele G was significantly enriched in this subset. Indeed, 8 subjects were homozygous for allele G and 40 were heterozygous AG. G allele frequency in these 77 males was 36%, which is higher than the 19% found in the individuals from Northwest Europe (CEU) used for the HapMap project (p<0.0001) and the 21% reported in 4300 European Americans from the National Heart, Lung, and Blood Institute (NHLBI) Exome Sequencing Project (p <0.0001). It is also higher than in the 26% found in the rest of the YD cohort (p=0.0029) which includes a large number of younger subjects recruited because of family history of hemochromatosis and of premenopausal women whose severity on the long term is difficult to assess. In conclusion, our study draws attention to two SNPs (rs884409 and rs11558492) in genes DCYTB and GNPAT, respectively, that are appear to worsen the phenotypic expression of YD patients. These results also support the hypothesis that GNPAT plays a still undiscovered role in iron homeostasis. 1. McLaren, C. E. et al. Exome sequencing in HFE C282Y homozygous men with extreme phenotypes identifies a GNPAT variant associated with severe iron overload. Hepatol. Baltim. Md (2015). doi:10.1002/hep.27851 The authors are grateful to the members of the "Centre de competence" on rare iron disorders of CHU de Montpellier Disclosures No relevant conflicts of interest to declare.

Histopathologic Study of Livers From Chronic Iron Loaded C282Y Mice: Model of Hereditary Hemochromatosis?

Abstract 5288 Introduction: The C282Y mouse model of hemochromatosis was developed to facilitate study of hereditary hemochromatosis in humans. We studied the liver histologic findings in mice that were iron loaded over at least 10 weeks to determine if the pathological changes in the mouse liver mirror the chronic iron overload condition observed in humans with genetic hemochromatosis. Methods: C282Y and wild-type mice (C57BL) were iron loaded over a minimum of 10 weeks using weekly intraperitoneal injections of 400 mg/kg of iron dextran. Animals were sacrificed at least 2 weeks after receiving their last iron injection. Hematoxylin and eosin, Masson's trichrome, and Perl's Prussian blue stains were used to assess the amount and distribution of liver iron and the presence or absence of inflammation and fibrosis. The slides were evaluated by a veterinary pathologist (MAS) without knowledge of genetic or iron load status. Specific histopathologic characteristics were evaluated and graded as the mean of 10 40x high power fields. These included: Hepatocyte iron distribution graded as either periportal or diffuse, amount of iron deposition in cells was graded for hepatocytes and Kupffer cells (graded 0–5), iron deposition in endothelial cells (graded 0–4), amount of fibrosis (graded 0–4), presence or absence of inflammation, periportal lymphocyte infiltration (graded 0–3), and hematopoiesis (present, present but mild, or absent). Livers from 24 mice (3 +/+, 9 +/Y and 12 Y/Y) were studied. Results: When compared with animals that were not iron loaded, animals iron loaded with at least 4 g/kg of iron had increased cellular deposition of iron. Of the two mice that were not iron loaded, the +/Y mouse had no increase in tissue iron and the Y/Y mouse had 2+ iron in hepatocytes, Kupffer cells, and endothelial cells. The highest two categories of cellular iron were observed in the four animals loaded with the highest doses of iron (5.6–6.8 g/kg): grade 4–5 for hepatocytes and Kupffer cells, and grade 3–4 for endothelial cells. Mice loaded with 4 g/kg of iron (7 Y/Y, 8 +/Y and 3 +/+) had similar but generally lower levels of iron deposition: grade 3–5 in hepatocytes, grade 3–4 in Kupffer cells, and grade 3 in endothelial cells. The three genotypes could not be differentiated based on the amount of iron deposition. However, the distribution of iron in hepatocytes of the 3 wild type animals (+/+) was clearly periportal which differed from the diffuse distribution pattern observed in the Y/Y and +/Y mice. One +/Y mouse had a mild periportal distribution of hepatocyte iron. Iron-loaded animals in all 3 genetic groups had grade 1 or 2 fibrosis in the periportal area and around the iron laden Kupffer cell aggregates. None of animals had bridging fibrosis or cirrhosis. In the iron-loaded Y/Y and +/Y mice, hematopoiesis was present in close association with intrasinusoidal iron laden Kupffer cell aggregates. In the non-iron loaded mice hematopoiesis was rare in the Y/Y animal, absent in the +/Y mouse. Hematopoiesis was present but mild in all 3 of the +/+ mice as well as in 1 mouse of each other genotype, and was present in the remaining 17 iron-loaded mice. No hepatocyte necrosis or regenerative nodules were detected in any mouse. Discussion: Histologic evaluation of the C282Y mice showed that the deposition of iron in the liver increased with increased iron loads. The distribution of iron between Kupffer cells and hepatocytes was similar in all 3 genetic groups. The periportal iron distribution observed in humans with genetic hemochromatosis was not seen in either the Y/Y or +/Y genotypes, but was observed in the iron-loaded wild type +/+ mice. Liver histology with chronic intraperitoneal iron loading in the C282Y C57BL mice is comparable to that observed in humans with secondary iron overload and not those observed in humans with hereditary hemochromatosis. Why the liver histology in the iron loaded C282Y knock-in mouse does not mirror the changes observed in human livers from patients with genetic hemochromatosis is not clear. It may be the dosing regimen or the strain of the mouse or that additional genetic or environmental factors are required. Disclosures: No relevant conflicts of interest to declare.