Hepcidin (encoded by HAMP), the peptide hormone regulating systemic iron homeostasis, controls iron flux to plasma and thus ensures adequate iron supply for erythropoiesis. Increased erythropoietic activity negatively regulates hepcidin synthesis by the hormone erythroferrone (ERFE) produced by erythroid precursors in response to erythropoietin (EPO). Here we evaluated iron metabolism in a mouse model of polycythemia with augmented erythropoiesis and low EPO levels bearing the gain-of-function polycythemia-causing human EPO receptor gene (mtHEPOR) (Divoky PNAS 2002). The mtHEPOR embryos developed polycythemia around embryonic day (ED) 17.5, followed by unexpected but transient correction of polycythemia in perinatal life (postnatal day 7 - PD7) and reappearance of a polycythemic phenotype at 3-6 weeks of age. (Human patients with the gain-of-function polycythemia-causing human EPOR mutations are also born non-polycythemic and then develop polycythemia within a few weeks of neonatal life). The mtHEPOR mice had reduced Epo levels, augmented EPOR signaling and prolonged activation of Stat5 in erythroblasts when compared to mEpoR mice (Divoky JMolMed 2016). To assess whether developmental changes of erythron in this mouse model correlate with changes in iron metabolism, we determined temporal changes of Erfe and Hamp expression and body iron stores.

We found that mtHEPOR mice have increased expression of Erfe at ED 17.5 (7.5-times) and PD7 (4-times) compared to mEpoR controls when normalized to a housekeeping gene beta-actin and independently to erythroid marker glycophorin A. Therefore Erfe in mtHEPOR mice would be increased both because of a higher number of erythroid precursors and because each precursor is producing more Erfe by increased transcription. This corresponds to reduced expression of Hamp in the liver of ED17.5 (4-times) and PD7 mtHEPOR mice (15-times) compared to mEpoR littermates. However, perinatally (PD1), Erfe expression dramatically decreased in mtHEPOR mice (7-times) concomitantly with increased Hamp expression (45-times), and both these factors reached expression levels detected in mEpoR at PD1. Expression of Erfe and Hamp between mtHEPOR and mEpoR mice was also not different in older mice (7 months of age), even though mtHEPOR mice were markedly polycythemic with persistent erythropoietic activity, markedly increased splenic erythropoiesis (the weight of spleens of control animals was 100-180 mg compared to 200-350 mg of mtHEPOR mice), and augmented EPOR-Stat5 signaling. However, mtHEPOR and mEpoR mice differed in transferrin saturation (TSAT) and tissue iron stores. The spleens of mtHEPOR mice had markedly increased iron deposits compared to mEpoR controls and stainable iron could only be detected in the livers of mtHEPOR mice but not in the livers of mEpoR controls. TSAT was elevated in mtHEPOR mice (77±3.2%) when compared to mEpoR (54.7±8.0%), consistent with the iron overload in mtHEPOR mice. Nevertheless, Hamp expression was comparable between mtHEPOR and mEpoR mice. This shows that Hamp expression was unexpectedly normal in respect to the presence of iron overload in mtHEPOR mice at 7 months of age.

With further aging (mice 14 months and older), Erfe continued to increase and Hamp continued to decrease in mEpoR mice. In contrast, there was a trend towards decreased Erfe expression while Hamp expression became significantly elevated (2-times) in mtHEPOR mice. These changes were accompanied by a 2.4-fold decline in TSAT in both genotypes, which suggests that aged mEpoR mice develop iron deficiency while mtHEPOR mice remain iron-replete likely due to mobilization of iron from high iron stores.

We show that the polycythemic phenotype, as well as Erfe and Hamp expression, in mtHEPOR mice undergoes dynamic changes during ontogenesis. We conclude that further research will be needed to understand the development of iron overload and changes in Erfe and Hamp expression in mtHEPOR mice characterized by erythroid hyperplasia and low Epo levels.

Acknowledgment: BK, VD, and MH were supported by the Czech Science Foundation (projects GA15-13732S and GA17-05988S), KONTAKTII project LH15223, and in part by project LTAUSA17142. BK and JS contributed equally to this work. JTP and MH act as equivalent co-senior authors.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.