Despite the recent discovery of new iron-related genes, gaps remain in our understanding of vertebrate iron homeostasis. For example, the role of copper in iron metabolism has puzzled investigators ever since 19th-century physicians found that copper salts, when given with iron, could overcome resistant cases of anemia.1 Equally mystifying is the mechanism by which iron is released from cells and tissues for use in erythropoiesis. A linkage between these problems was first described in 1931 by Cooke and Spilles, who showed that administration of copper to iron-deficient rats mobilized spleen iron stores.1 Later, Earl Frieden and coworkers proposed a biochemical mechanism to explain the role of copper in tissue iron release.1 According to their model, the ferroxidase activity of ceruloplasmin converts ferrous iron to ferric iron for high-affinity binding to apotransferrin. The formation of iron-loaded transferrin creates, from the perspective of the cell, a negative, free ferrous ion concentration gradient, thereby enhancing iron release.
In this issue of Blood, Harris and colleagues (page 4672) examine the mechanism by which ceruloplasmin stimulates tissue iron release in vivo. They take advantage of recent studies in yeast, mice, and humans that have provided critical insights and technologies. These authors have already made several seminal contributions. For example, they have described people with ceruloplasmin gene defects who have low serum iron and transferrin saturation but debilitating tissue iron overload.2 This finding provides evidence for a role of ceruloplasmin in tissue iron release, a conclusion supported by kinetic studies in mice with targeted ceruloplasmin gene disruption.3 An additional molecular clue comes from the discovery that Fet3p, a membrane-bound protein critical for high-affinity iron uptake in Saccharomyces cerevisiae, is, like ceruloplasmin, a multicopper protein with ferroxidase activity.4
Here, the authors show that injection of recombinant Fet3p into aceruloplasminemic mice restores serum iron and transferrin saturation. The results are remarkable for both the speed and magnitude of the effect; restoration to wild-type levels is essentially complete by 1 hour after injection. More importantly, the results have significant implications regarding in vivo mechanisms of ceruloplasmin-stimulated iron flux. At least 3 mechanisms can be invoked: ceruloplasmin interacts with cell surface iron transporters (or signaling receptors) to increase iron flux; ceruloplasmin interacts with transferrin to increase iron binding; or ferroxidase activity itself is rate-limiting in iron transport, and specific ceruloplasmin-protein interactions are not required. The authors exploit the near-total lack of sequence homology between ceruloplasmin and Fet3p (except for the few amino acids coordinating interior copper binding) to deduce that the third mechanism is operative. They conclude that physical interactions of ferroxidase are not essential for iron release but rather that the process is diffusion-limited and that ceruloplasmin, by accelerating the formation of ferric ion-transferrin complex, may reduce the return of ferrous iron to cells. One should recognize a caveat to this argument: namely, that there may be structural (rather than sequence) similarity between Fet3p and ceruloplasmin that facilitates a specific interaction between Fet3p and a membrane receptor or transporter. The authors' conclusion is consistent with recent cell culture studies in which iron release was shown to be nonsaturable with respect to ceruloplasmin concentration, a finding inconsistent with a specific-binding mechanism.5 In contrast, binding of ceruloplasmin to the cell surface iron transporter, iron-regulated gene 1 (IREG1)/metal transport protein 1 (MTP1)/ferroportin, has been reported in astrocytes.6 The relevance of this result to systemic iron release is uncertain, since astrocyte ceruloplasmin is uniquely glycosylphosphatidylinositol anchored and since the relationship between this interaction and iron transport has not yet been shown. The work of Harris and colleagues has striking implications regarding protein function. As proposed by the authors, the results indicate an evolutionary conservation of the function of multicopper oxidases in iron metabolism. Furthermore, the results suggest that the oxidases determine the amount but not the direction of iron flow. This concept is most emphatically illustrated by the fact that one protein, Fet3p, can drive either iron uptake in yeast or iron release in mice.