Transferrin, ferritin and the race to recycle heme

Iron routing comes before iron withholding

Most host iron is useful precisely because it is unavailable as a freely exchangeable nutrient. The central question is where an iron atom sits, which ligand holds it, and which cell can retrieve it. Transferrin transports ferric iron through extracellular fluid; ferritin stores an intracellular mineral reserve; hemoglobin carries heme inside red cells. These pools are connected, but an assay of total iron collapses distinctions that determine microbial access. ^{[1] [5]}

During normal red-cell turnover, macrophages recover iron and return it to the circulation for reuse. This is indispensable homeostasis. Infection can redirect the same machinery, while tissue damage can expose pools that were previously inaccessible. Calling every binding event an evolved antimicrobial defense therefore skips a step: one needs evidence of microbial restriction, host protection, or pathogen-driven adaptation. These are related outcomes, but they are experimentally separable. ^{[1] [2] [4]}

Follow iron chemistry and microbial access in the iron profile →

Transferrin is a carrier under evolutionary attack

Barber and Elde supplied unusually direct evidence that a transport protein also participates in an evolutionary conflict. They compared transferrin across primates, mapped rapidly evolving residues to the bacterial TbpA-binding surface, and tested substitutions experimentally. Changing individual transferrin residues altered recognition by bacterial receptors. The combined evolutionary and binding evidence is stronger than simply observing that transferrin binds iron. ^[2]

The result was also selective. The human C2 transferrin variant resisted recognition by some Haemophilus influenzae TbpA variants, whereas tested Neisseria receptors retained recognition. This was recombinant-protein work, not a demonstration that C2 carriers cannot become infected. Its significance is conceptual: a host can change the accessibility of an iron package while preserving the requirement to transport iron. A pathogen can answer by changing the protein that opens the package. ^[2]

The remaining physiological question is whether these binding differences preserve transferrin delivery to host cells. Host-receptor recognition, iron release, and cellular uptake would need separate assessment before a variant could be described as beneficial overall. A change that also compromises host delivery would introduce a tradeoff. The published bacterial binding data establish differences in recognition; they do not establish clinical resistance or the net balance between iron transport and defense. ^[2]

Ferritin protection depends on turnover and context

Ferritin is a reversible reserve rather than a permanent disposal route. Mancias and colleagues identified NCOA4 as a cargo receptor that helps deliver ferritin for lysosomal degradation, a process called ferritinophagy. Depleting NCOA4 impaired ferritin turnover and changed iron-responsive proteins in cultured human cells. More stored protein can therefore coexist with less usable iron. These experiments established a mobilization mechanism; they did not determine whether a particular intracellular pathogen receives the released iron. ^[5]

Haschka and colleagues tested ferritin H more directly during Salmonella infection. Myeloid Fth deletion increased macrophage labile iron and disturbed storage. After experimental intravenous iron loading, deficient mice developed greater bacterial burdens, hyperinflammation and shorter survival. Crucially, the survival difference was absent without additional iron loading. The result supports conditional protection by ferritin buffering, not an unlimited rule that increasing ferritin always suppresses infection. ^[8]

A complementary perturbation shows why storage cannot be interpreted in isolation. Nairz and colleagues removed macrophage iron-regulatory proteins IRP1 and IRP2, producing excess ferritin and greater susceptibility to Salmonella. Radioiron tracing showed loss of label from the enlarged ferritin pool while bacterial label increased, supporting an exploitable intracellular reserve. The route of ferritin-iron mobilization remained unresolved. Because IRP loss also altered lipocalin-2 and other responses, this is not the inverse of deleting ferritin H: it changes a regulatory network, not just storage capacity. Together, the studies demand a flux-based explanation. ^{[9] [8]}

Storage, release and export must consequently be measured together. A macrophage can contain substantial iron while making little available to a vacuolar bacterium, or it can lose buffering capacity while containing less stored iron overall. This is why “intracellular iron” is still too coarse a category: cytosol, lysosome and pathogen-containing compartment are different destinations. The practical research question is which flux changes before bacterial growth changes. ^{[5] [8]}

See how hepcidin and ferroportin alter the export side of this system →

Haptoglobin and hemopexin route different cargo

When red cells rupture within vessels, haptoglobin captures released hemoglobin. Kristiansen and colleagues identified CD163 as a receptor mediating uptake of the complex in transfected and myelomonocytic cells. The important advance was connecting extracellular capture to cellular internalization. Binding a dangerous protein and removing that complex from plasma are separate steps; neither means that its iron has left the body. ^[6]

Hemopexin handles heme released from hemoproteins. Hvidberg and colleagues identified LRP/CD91 as a receptor for hemopexin–heme complexes, showing uptake in receptor-expressing cells, lysosomal processing and induction of heme oxygenase-1 transcription in cultured monocytes. This provides a route from extracellular scavenging to intracellular metabolism. It is not itself an infection challenge or evidence that every bacterium is denied the bound heme. ^[7]

This sequence gives a more precise meaning to protection. Scavenging can reduce exposure to damaging extracellular hemoglobin or heme; intracellular processing can preserve host cells; iron routing can influence microbial nutrition. An experiment showing the first effect cannot assume the third. In evolutionary terms, the homeostatic necessity of these pathways does not by itself tell us which component was selected specifically to resist infection. ^{[6] [7] [4]}

Microbes can exploit the carrier itself

Zambolin and colleagues showed how H. influenzae HxuA disrupts hemopexin’s hold on heme. Structural analysis and mutant experiments connected a mobile HxuA loop to heme release. A loop-deletion protein could still bind hemopexin yet failed to support acquisition from it in the reconstructed system. That separation is decisive: binding the host protein and extracting its cargo are different functions. The structural work used rabbit hemopexin, and key growth tests used engineered heme-auxotrophic E. coli rather than infected people. ^[3]

The lesson is not that hemopexin is ineffective. It is that protection depends on the organism’s equipment. In Torres and colleagues’ experiments, Staphylococcus aureus Newman used hemoglobin but did not grow on hemopexin under the tested conditions. Isotope tracing distinguished hemoglobin-derived iron from transferrin-derived iron and showed preferential acquisition from hemoglobin, with the balance shifting during growth. Disrupting IsdB impaired hemoglobin use and reduced bacterial burden in a mouse infection model. ^[10]

These studies make a common assumption testable: “bound” cannot be used as a synonym for “unavailable.” The relevant comparison is access by the host versus access by a specified microbial strain. Even a high-affinity host carrier may present a recognizable surface for an extraction apparatus. Conversely, success against one iron pool does not demonstrate access to another. A receptor inventory is therefore a hypothesis about nutrition until uptake and growth are measured. ^{[3] [10]}

Compare these extraction strategies with lactoferrin withholding and microbial countermeasures →

Readouts must separate nutrition from tissue protection

Seixas and colleagues demonstrated an especially important distinction in experimental malaria. Heme oxygenase-1 protected mice against liver injury and lethality during Plasmodium chabaudi chabaudi infection without a corresponding reduction in parasitemia in the principal comparisons. This is disease tolerance: surviving a given pathogen burden more successfully. It differs from resistance, which lowers that burden. A heme-handling intervention could therefore succeed biologically even when parasite counts remain unchanged. ^[4]

Clinical ferritin is similarly informative without being a complete mechanistic readout. WHO guidance treats inflammation as a material influence on serum ferritin and advises against using ferritin alone to identify iron overload. For this field’s research questions, a ferritin rise is not proof that bacteria lost access to iron, nor does a fall prove that host stores were safely removed. Inflammatory context and the other measured compartments determine what the change can mean. ^[11]

A useful panel follows a causal sequence: identify the ligand carrying iron, quantify its movement into host and microbial fractions, measure a validated labile-iron response, and then measure pathogen abundance and host injury separately. Time resolution matters. A late change in ferritin after extensive tissue damage cannot establish which earlier flux drove infection. Source-specific tracers provide information that total elemental analysis cannot supply on its own. ^{[10] [11]}

A next experiment: capture, clearance, and host injury

The useful result would be a host-side map of destinations and consequences: extracellular capture, cellular uptake, storage, export, and injury. It would clarify which meaning of protection the hemopexin pathway supports. Published microbial extraction mechanisms would remain a reason to avoid inferring nutritional exclusion from that map. Establishing safer host handling of heme and establishing reduced microbial access are distinct scientific conclusions. ^{[7] [3] [4]}

Place iron routing within the wider host metal defense series →