Serum Erythroferrone Levels in a Large General Population: Towards a Better Understanding of Connections between Erythropoiesis and Iron Homeostasis

Introduction: Iron is essential for various biological functions [ Ganz T., Physiol Rev 2013], especially for erythropoiesis which is the most iron-consumer process daily. Hepcidin is the key hormone of iron homeostasis, preventing both iron deficiency and excess, by modulating iron absorption, recycling, and mobilization [ Camaschella C. et al, Haematologica 2020]. Erythroferrone (ERFE) constitutes the principal erythroid regulator of hepcidin: in the presence of enhanced erythropoiesis, expanding erythroblasts produce ERFE which, by impeding bone morphogenetic proteins (BMPs) signaling in hepatocytes, suppress hepcidin, in turn increasing iron availability for the bone marrow [ Ganz T., Free Rad Biol Med 2019]. Increased ERFE levels have been demonstrated in expanded/ineffective erythropoiesis (IE) (e.g., in thalassemia intermedia), where chronic ERFE-mediated hepcidin suppression may cause progressive iron accumulation even in subjects not regularly transfused. In view of this, ERFE has been proposed as a potential biomarker of IE and an appealing target for treating iron overload deriving from IE [ Srole D.N. and Ganz T., J Cell Physiol 2020]. However, studies that fully elucidate the role of ERFE in the general population under physiological conditions are still lacking. Our project evaluated age- and sex-related variations in ERFE levels, and their correlations with hematological and iron status parameters, in a large subpopulation from the Cooperative Health Research In South Tyrol (CHRIS) study.

Subjects and methods: The CHRIS study is a population-based study carried out in South Tyrol (Northern Italy) aimed at investigating acquired and genetic basis of common chronic conditions associated with aging [ Pattaro C. et al, J Transl Med 2015]. Blood samples were tested for several biochemical and molecular parameters, including those related to iron status, such as ferritin, transferrin saturation (TSat), soluble transferrin receptor (sTfR), hepcidin, and C282Y and H63D mutations. Serum ERFE was measured in stored aliquots by the Human Erythroferrone IETM ELISA kit (Intrinsic Lifesciences - The BioIron Company TM, USA), according to manufacturer instructions.

Results: We measured ERFE in 3688 subjects aged >18 years (55% female). A total of 777 subjects (21% of the whole population; 57% females) had ERFE concentrations below the lower limit of detection for our method (we assigned them an arbitrary value of 0.16 for statistical analyses). Figure 1 illustrates age- and sex-dependent variations of ferritin, sTfR, hepcidin, and ERFE in the whole population. After stratification by hemoglobin (Hb) levels, we observed that ERFE was significantly higher in anemic (i.e., Hb<12 g/dl in females and <13 g/dl in males; n=100) than in non-anemic subjects (respectively 3.74 and 0.65 ng/ml; p<0.0001). In anemic subjects, ERFE showed a strong negative correlation with Hb (r = -0.5; p < 0.001), TSat (r = -0.58; p < 0.001), ferritin (r = -0.48; p < 0.001) and hepcidin (r = -0.32; p = 0.001), and a positive correlation with sTfR (r = 0.58; p < 0.001). Moreover, ERFE was higher in iron-deficient (ID) subjects (i.e., ferritin <30 mg/l; n=773) than in non-ID subjects (0.92 vs. 0.63 ng/ml; p<0.0001). Although ID subjects had lower Hb levels than non-ID subjects (13.7 vs 15 g/dl; p<0.0001), Hb levels were broadly within the normal range.To better explore the relationship between ERFE and iron status, we divided the population into 4 subgroups, according to sTfR and ERFE levels. Considering the skewed distribution of the variables, we used sTfR above or below the 3° tertile (1316 ng/ml), in order to identify subjects having higher or lower iron needs, respectively. Analogously, we further divided the population in subjects with ERFE above or below the 3° tertile (1.21 ng/ml). As illustrated in Figure 2, after stratification, we observed that hepcidin reduced, in subjects with high ERFE levels, when iron requirement was high (0.94 vs 2.10 nM; p < 0.0001), but it didn't when iron requirement was low (3.35 vs. 3.05 nM; p = 0.72), even in presence of normal Hb levels (Hb 14.3 g/dl).

Discussion: our data confirmed an association among ERFE, Hb and iron status parameters under conditions of anemia (when erythropoiesis is stimulated); moreover, if confirmed in further studies, it could suggest an effect of ERFE on hepcidin regulation could be linked to depletion of iron stores, even before anemia develops.