Non–transferrin-bound iron takes the driver’s seat

In this issue of Blood, Charlebois et al¹ report that non–transferrin-bound iron (NTBI) is the primary driver of bone morphogenetic protein 6 (BMP6) expression in liver sinusoidal endothelial cells (LSECs) during iron overload. This finding is important because LSEC-derived BMP6 prompts the liver to produce hepcidin, the chief iron-regulatory hormone that regulates body iron balance.

Otto von Bismarck, the “Iron Chancellor” of the German Empire in the late 19th century, argued that great questions of national policy are settled by iron and blood. One can make a similar argument for present-day research in iron biology, in which a great and unresolved question will be settled by iron. In essence, the question is how the body “senses” iron status so that it can adapt to absorb more iron when needed but avoid accumulating too much of the metal, which can be toxic in excess. Such regulation is essential for body iron balance because humans cannot excrete excess iron. An important advance in recent years has been the identification of LSECs as the site of iron sensing.² LSECs respond to iron loading by increasing the expression and secretion of BMP6,^3,4 which activates in neighboring hepatocytes a signaling pathway that induces the expression of hepcidin,⁵ the hormone that controls how much iron the intestine absorbs. However, the form of iron taken up by LSECs that triggers Bmp6 expression in vivo and the molecular mechanisms involved have not been well defined.

The 2 most plausible candidates for conveying the iron signal to LSECs are transferrin-bound iron and NTBI. In normal blood plasma, iron circulates nearly exclusively as transferrin-bound iron (ie, holotransferrin), which cells take up via transferrin receptor 1 (TFR1)-mediated endocytosis. In iron overload conditions, plasma iron increases to levels that exceed the iron-carrying capacity of transferrin, giving rise to NTBI, a poorly defined, heterogenous, and variable mixture that includes ferric citrate and high-mass iron aggregates.^6,7 Usually undetectable in normal healthy individuals, plasma NTBI becomes measurable when transferrin saturations surpass 70%, such as in the iron overload disorders hereditary hemochromatosis and thalassemia major. Cells take up NTBI via divalent metal-ion transporters such as ZIP14, ZIP8, and DMT1.⁶ Although previous studies have shown that either holo-transferrin or NTBI (as ferric ammonium citrate) can load primary mouse liver endothelial cell cultures with iron and induce Bmp6 expression,^3,4 how these iron sources contribute to LSEC BMP6 production in vivo requires clarification. Using mouse models and single-cell transcriptomics, Charlebois et al conclude that NTBI is the main regulator of LSEC BMP production during iron overload.

To define the role of LSEC TFR1 in the iron-dependent regulation of Bmp6 expression, the authors generated mice with endothelial-specific inactivation of the TFR1-encoding Tfrc gene. They found that mice lacking endothelial TFR1 display no alterations in systemic or tissue iron levels and express normal amounts of BMP6 and hepcidin, indicating that endothelial TFR1 does not play a major role in iron homeostasis. However, when fed an iron-deficient diet, mice lacking endothelial TFR1 displayed reduced expression of hepatic BMP6 and hepcidin relative to liver iron content, suggesting a minor requirement for endothelial TFR1 during iron deficiency. The need for TFR1 under these conditions makes sense because cells respond to iron deficiency by upregulating TFR1 levels to acquire more iron from holo-transferrin. Of note, a nearly identical iron phenotype and iron-deficiency response was recently reported by Fisher et al,³ who studied mice lacking endothelial TFR1 generated by using a different Cre mouse line (ie, Stab2-Cre vs Tek-Cre).

In their study, Charlebois et al additionally demonstrated that feeding mice a high-iron diet, which elevated transferrin saturations to more than 90%, increased hepatic expression of Bmp6 and hepcidin similarly between wild-type mice and those lacking endothelial TFR1. Moreover, plasma NTBI concentrations of mice of either genotype correlated positively and significantly with hepatic Bmp6 and hepcidin messenger RNA (mRNA) levels. Single-cell transcriptomic analyses of liver cell types from wild-type mice further revealed a more robust induction of Bmp6 mRNA in LSECs following dietary iron loading (and NTBI formation) than from a single intravenous bolus of exogenous holo-transferrin. Together, these data imply that LSEC uptake of NTBI, rather than transferrin-bound iron via TFR1, drives Bmp6 and hepcidin expression in response to iron loading (see figure). Future studies need to define temporal and dose-response effects of plasma NTBI (eg, exogenously administered NTBI) on Bmp6 induction and hepcidin expression in vivo.

But how do LSECs take up NTBI? The most well-characterized NTBI transporter, ZIP14 (SLC39A14), which mediates NTBI uptake by hepatocytes, appears dispensable as previous studies showed that Slc39a14 knockout mice efficiently up-regulate hepatic Bmp6 and hepcidin expression in response to short- or long-term dietary iron loading or genetic iron overload.⁸ Seemingly consistent with this idea, the authors’ single-cell transcriptomic analyses of LSECs from wild-type mice found that Slc39a14 mRNA levels did not increase in response to a high-iron diet. The transcriptomic analyses did reveal, however, that the high-iron diet strongly induced the expression of Slc39a8 encoding ZIP8, which like ZIP14, can transport NTBI, at least in cultured cells.⁹ This finding, along with the observation that LSECs abundantly express Slc39a8,¹⁰ identifies ZIP8 as a lead candidate for NTBI uptake by LSECs.

If iron acquired from NTBI per se proves to be the dominant signal for BMP6 induction by LSECs in vivo, follow-up studies will need to identify which chemical species of NTBI mediates the effect. This new knowledge, combined with identifying the LSEC NTBI transporter(s), not only will add to the repertoire of possible therapeutic targets for treating iron overload disorders but will also help to fill the gap in our understanding of how iron regulates its own homeostasis.

Conflict-of-interest disclosure: M.D.K. has consulted for Pharmavite.