ABCs of erythroid mitochondrial iron uptake

ATP-binding cassette (ABC) transporter proteins are present in all known species, serving a variety of important functions that generally involve translocation of small molecules across cellular membranes (1). In this issue of PNAS, Chen et al. (2) describe a novel role for one member of the ABC transporter superfamily: Abcb10 stabilizes a mitochondrial iron transporter from an entirely different protein family to optimize hemoglobin production by erythroid precursors.

Fig. 1.

A model of the Mfrn1 and Abcd10 interaction in the differentiating erythroid precursors. (A) Mrfn1 is relatively unstable in erythroid progenitor cells in the absence of Abcd10, limiting mitochondrial iron transfer. (B) During erythroid differentiation, Mfrn1 is stabilized by Abcd10 in the inner mitochondrial membrane to increase iron transfer. (C) After heme production is complete, heme acts to decrease Abcd10 protein levels, leading to turnover of Mfrn1.

Hemoglobin, the most abundant heme-containing protein in mammals, gives red blood cells their color. Heme production occurs in all cells, but high-level production in early erythroid cells is achieved with the help of several erythroid-specific molecules, including a specialized form of the first enzyme of heme biosynthesis and the product of an erythroid-restricted splice isoform of a later enzyme in the pathway (3, 4). Coordination of heme synthesis with translation of globin proteins is important to avoid toxic accumulation of hemoglobin precursors. Heme synthesis begins and ends in the mitochondria, but several intermediate steps take place in the cytoplasm, requiring transport of biochemical precursors across the mitochondrial membrane. The ultimate step in heme biosynthesis, insertion of an iron atom into protoporphyrin IX, is catalyzed by mitochondrial ferrochelatase. Until recently, it was not clear how iron was delivered to that enzyme.

In 2006 Paw and colleagues (5) identified the gene mutated in anemic frascati zebrafish and showed that its product, mitoferrin (Mfrn1, also known as Slc25a37), transports iron across the mitochondrial membrane. In collaboration with Kaplan's group, they noted that Mfrn1 protein was stabilized as erythroid precursors underwent differentiation, consistent with a key role in mitochondrial iron acquisition (6). Now, they show that Mfrn1 stabilization is accomplished through interaction with another protein that is highly restricted to erythroid mitochondria, Abcb10 (also known as Abc-me) (2).

Chen et al. (2) have determined which portion of Mfrn1 physically interacts with Abcb10, but how it prevents Mfrn1 turnover remains unknown. Abcb10 does not appear to be an iron transporter itself, and its substrate, if any, has not been identified (7). Abcb10 is down-regulated in response to treatment of cells with heme, suggesting a mechanism for indirectly decreasing Mfrn1 activity late in erythroid differentiation. Additional work will be needed to determine which of the canonical ABC transporter activities (ABC hydrolysis, transmembrane translocation) Abcb10 possesses, and how these relate to its role in stabilizing Mfrn1.

Mfrn1 and Abcb10 can both homodimerize (2, 8), but Chen et al. (2) detect Mfrn1 in a complex that is much larger than would be expected for a heterotetramer, suggesting that other proteins may also participate in a multicomponent protein machine. The experiments reported here provide no direct clues as to what the other components might be, but it is intriguing that coexpression of Mfrn1 and Abcb10 in undifferentiated mouse erythroleukemia cells leads to increased mitochondrial iron import but not increased heme biosynthesis. It may simply be, as the authors suggest, that protoporphyrin precursors are insufficient. An alternative explanation, however, might be that ferrochelatase itself is a part of the complex and is limiting. Ferrochelatase is also an integral protein of the inner mitochondrial membrane (9, 10). It would be particularly efficient for it to be physically associated with the transporter that brings iron into the mitochondrion. As such, it would be a very logical candidate for the next component of the Mfrn1/Abcb10 complex.

Zebrafish lacking functional Mfrn1 develop anemia characterized by an arrest in erythroid differentiation (5). If the major function of Abcb10 were to stabilize Mfrn1, as proposed by Chen et al. (2), one would predict that animals lacking Abcb10 would also be anemic. A recent abstract reporting unpublished work suggests that this is the case—mice deficient in Abcb10 do, in fact, develop an embryonic-lethal, hypoplastic anemia consistent with a defect in heme biosynthesis.*

The Mfrn1/Abcb10 complex allows erythroid precursors to accelerate their production of heme by efficiently providing iron to ferrochelatase. Abcb10 is not abundant, however, in other tissues where heme proteins have major functions and cells express high levels of ferrochelatase, such as muscle, liver, and adrenal gland. Will homologous or analogous proteins serve a similar function?

Identification of the Mfrn1/Abcb10 mitochondrial iron transport machine is an important step toward fully understanding the molecular regulation of erythroid iron utilization. Until now, very little information was available about the path taken by iron between the endosome, where it is internalized bound to transferrin and the transferrin receptor, and the mitochondrion, where it is incorporated into protoporphyrin. Ponka and colleagues (11) have proposed that iron is transferred between endosomes and mitochondria through direct contact and interaction of the organelles, but the topological and molecular features of this model have not been described in detail. It should now be feasible to investigate the intracellular iron shuttling mechanism more directly.