The highly variable clinical phenotype observed in patients homozygous for the C282Y mutation of the hereditary hemochromatosis gene (HFE) is likely due to the influence of non-HFE modifier genes. The primary functional abnormality causing iron overload in hemochromatosis is hyper-absorption of dietary iron. We found that iron absorption in inbred mice varies in a strain-specific manner, as does the pattern of iron distribution to the liver and spleen. A/J mice absorbed approximately twice the amount of 59Fe delivered by gavage compared to the C57BL/6 strain. Genetic comparisons between A/J and C57BL/6 were facilitated by the availability of consomic chromosome substitution strains (CSS). Each CSS has an individual chromosome pair from A/J on an otherwise C57BL/6J background. We found that iron absorption and iron content in liver and in spleen were continuous variables suggesting that each trait is under multigenic control. No trait co-segregated among the CSS. Chromosome 5 from A/J, however, imparted the highest iron absorption phenotype and multiple CSS had absorption levels equivalent to A/J. Chromosomes 9 and X were associated with high spleen iron content. These data suggest that multiple genes contribute to the regulation of iron absorption and that individual organ iron phenotypes are independently regulated.

Mammalian iron homeostasis is maintained at the level of absorption from the gut lumen. Iron absorption is known to respond to changes in iron stores, erythropoiesis, and hypoxia. Normal individuals absorb approximately 1 mg of dietary iron daily. In mild to moderate iron deficiency, this value increases to 3-5 mg/day. Estimates of iron absorption during severe iron deficiency or anemia range from 30-40 mg/day. Hypoxia response appears to be intermediate. We asked how mice respond to differences in dietary iron content. Groups of 8 mice were weaned onto diets containing 2, 35, 120, 350, and 2,000 mg/kg carbonyl iron for 4 wk, following which iron stores, hematological parameters, iron absorption, and expression of iron-related genes were compared. Mean hepatic iron content was 52.7 ± 3.7 μg/g (wet wt) in mice fed the 2 mg/kg diet and increased to 560 ± 23.7 μg/g in mice fed the 2,000 mg/kg diet. There was no difference in hepatic iron in mice fed 35-350 mg/kg diets and averaged 110 ± 3 μg/g. Mean spleen iron content was 132 ± 12.2 μg/g (wet wt) in mice maintained on 2 mg/kg and increased to 598 ± 49 μg/g in mice fed the 2,000 mg/kg diet. There was no difference in spleen iron in mice maintained on intermediate iron diets (mean 359 ± 10 μg/g). These data indicate that iron stores remained constant over a 10-fold range in dietary iron content and changed only at the extremes. Erythropoietic demand did not change over the entire range of dietary iron as no differences between groups were noted in hematocrit, hemoglobin, and MCV. Iron absorption was measured as percent of a measured dose of 59 Fe (5 μg total) remaining in the carcass (minus the GI tract) 24 h after administration by gavage. Absorption was inversely proportional to dietary iron content. Mean absorption was 86% ± 4, 42% ± 3, 26% ± 7, 19% ± 4 and 6% ± 1 on the 2, 35, 120, 350, and 2,000 mg/kg iron diets respectively. Real-time PCR was used to measure liver hepcidin mRNA. Hepcidin expression was 20-fold greater in mice on the 2,000 mg/kg diet than in mice on the 2 mg/kg diet (3,900 ± 1,021 vs 198 ± 47 copies/actin copy). Hepcidin expression did not differ in mice on intermediate diets (745 ± 147 copies). These data indicate that under conditions where iron stores are not changing and there is no evidence of altered erythropoiesis, iron absorption remained exquisitely sensitive to dietary iron content. This result suggests that over a broad range of dietary iron content, normal iron homeostasis is regulated by a factor(s) intrinsic to the enterocyte and not by “downstream” effects of iron stores and erythropoiesis.

Dietary iron absorption by enterocytes is mediated by a ferrous transporter (DMT1) and possibly a ferric reductase (Cbyrd1). The role of Cbyrd1 is uncertain, as a knockout mouse has no defect in absorption (Gunshin et al. Blood, 2005, June 16, Epub). Transfer of iron to plasma is mediated by ferroportin (FPN). FPN’s residence on the basolateral membrane is regulated by hepcidin (Nemeth et al. Science 2004, 306:2090–3). Iron absorption responds to erythropoiesis, hypoxia and iron stores. A dietary iron content of 120 mg/kg consumed will maintain hepatic iron stores near that of mice found in the wild. Mice will grow and breed given diets containing 35 mg/kg. Commercial mouse chow iron content ranges from 200–350 mg/kg. We studied the effects of diets containing 2, 35, 120, 350 and 2000 mg iron/kg on iron absorption, liver and spleen iron content and transcriptional levels of DMT1, FPN and hepcidin in A/J mice. Mice were weaned at 3 weeks of age and groups of 8 animals were placed on one of the 5 diets for 4 weeks. No differences between groups were noted in hematocrit, hemoglobin and MCV. Mean hepatic iron content was 52.7 ±3.7 ug/g (wet wt) in mice fed the 2 mg/kg diet. Mean hepatic iron content was 560 ±23.7 ug/g in mice fed the 2000 mg/kg diet. There was no difference in hepatic iron content in mice fed intermediate iron diets (35–350 mg/kg). Mean hepatic iron concentration in these groups was 110 ±3 ug/g. Mean spleen iron content was 132 ±12.2 ug/g (wet wt) in mice maintained on 2 mg/kg. Mean spleen iron content was 598 ±49 ug/g in mice fed the 2000 mg/kg diet. There was no difference in spleen iron content in mice maintained on intermediate iron diets (mean 359 ±10 ug/g). These data indicate that mice maintain constant levels of hepatic and splenic iron over a ten-fold range in dietary iron content. Iron absorption was measured as percent of a measured dose of 59Fe (5 ug total) remaining in the carcass (minus the GI tract) 24 h after administration by gavage. Absorption was inversely proportional to dietary iron content. Mean absorption was 86% ±4 in the group on the 2 mg/kg diet, 42% ±3 on the 35 mg/kg diet, 26% ±7 on the 120 mg/kg diet, 19% ±4 on the 350 mg/kg diet and 6% ±1 on the 2000 mg/kg diet. Transcript levels of hepcidin, DMT1 and FPN were measured by real-time PCR and normalized to beta actin mRNA. Liver hepcidin expression was 20-fold greater in mice on the 2000 mg/kg diet than in mice on the 2 mg/kg diet (3900 ±1021 copies/actin copy vs. 198 ±47). Hepcidin expression did not differ in mice on intermediate diets (745 ±147 copies). Enterocytes were isolated from everted gut explants by elution in EDTA. Transcript levels of enterocyte DMT1 and FPN were 4566 ±SEM and 236 ±SEM copies respectively in mice on the 2 mg/kg diet. No detectable transcripts were found in mice on the 2000 mg/kg diet. Enterocyte transcript levels for DMT1 and FPN were no different in groups on intermediate iron diets (17 ±2 copies and 20 ±9 copies respectively). These data indicate that tissue iron content, hepcidin, DMT1 and FPN remain constant over a ten-fold range in dietary iron and only vary at extremes, while iron absorption is inversely proportional to dietary iron. The data also suggest that dietary iron, within defined limits, regulates iron absorption by a mechanism intrinsic to the enterocyte.