Gene: TFR2 · UniProt: Q9UP52 · Organism: Homo sapiens (human)
Protein family (InterPro/UniProt): Peptidase M28, subfamily M28B (a non-catalytic, "pseudo-peptidase" receptor branch); PA domain; TfR-like dimerization domain.
The gene symbol TFR2 unambiguously matches the UniProt record Q9UP52: "Transferrin receptor protein 2 (TfR2)," human. The literature is abundant and internally consistent for exactly this protein: a type II single-pass transmembrane receptor, homolog of transferrin receptor 1 (TFR1/TFRC), predominantly expressed in hepatocytes and erythroid precursors, functioning in systemic iron homeostasis. The domain architecture provided (peptidase M28B + PA domain + TfR-like dimerization domain) is precisely the architecture of the transferrin-receptor family. Importantly, although TFR2 is classified in peptidase family M28, it is a non-catalytic member — no proteolytic activity has been described; the protease-like fold instead serves as a ligand-binding/receptor scaffold, exactly as for its homolog TFR1. No gene-symbol ambiguity was encountered, and all retrieved literature concerns the correct protein.
TFR2 is not an enzyme or a transporter of iron in the classical sense — it is a cell-surface iron-sensing receptor. Its primary molecular function is to bind diferric (holo-) transferrin and, in response, to positively regulate expression of the iron-master-hormone hepcidin in hepatocytes, thereby controlling systemic iron balance. In erythroid cells it has a second role: it associates with the erythropoietin receptor (EPOR) and adjusts red-cell production to iron availability. Loss of TFR2 causes hereditary hemochromatosis type 3 (HFE3), a systemic iron-overload disease.
TFR2 is a receptor for diferric (iron-loaded) transferrin (holo-Tf), not a catalytic enzyme. Its specificity is for the iron-loaded form:
- Adding diferric Tf (but not apo-Tf and not non-Tf-bound iron) to hepatoma cells raises TFR2 protein in a dose- and time-dependent, reversible, hepatocyte-specific manner. The increase is post-transcriptional, reflecting an increased protein half-life (protein stabilization), not more mRNA [Johnson & Enns, 2004, PMID 15319290].
- Ligand binding requires an RGD (Arg-Gly-Asp) motif; the G679A mutant cannot bind diferric Tf and is no longer regulated by ligand [Pagani et al., 2015, PMID 25637053].
- N-linked glycosylation (Asn 240/339/754) is not required for surface trafficking or Tf binding but is required for holo-Tf-induced stabilization and thus for the iron-sensing readout [Zhao & Enns, 2013, PMID 23556518].
Thus TFR2 works as a sensor of the diferric-transferrin concentration in blood: as transferrin saturation rises, holo-Tf stabilizes surface TFR2 (partly by blocking its iron-dependent cleavage/shedding), converting an extracellular iron signal into a receptor-abundance signal [PMID 15319290; PMID 25637053].
Once engaged by holo-Tf, TFR2 drives transcription of hepcidin (HAMP) through a receptor complex and a defined kinase cascade:
- Membrane complex: HFE, TFR2, and the BMP co-receptor hemojuvelin (HJV) assemble into a multiprotein plasma-membrane complex on hepatocytes; TFR2 residues 120–139 are required to bind both HFE and HJV. TFR2 and HJV each compete with TFR1 for HFE, so the balance of these receptors sets iron sensing [D'Alessio et al., 2012, PMID 22728873].
- Kinase cascade: Holo-Tf activation of TFR2 stimulates ERK1/2 (MAPK) phosphorylation, which cross-talks with and boosts the canonical BMP6–SMAD1/5/8 pathway; the ERK inhibitor U0126 blunts holo-Tf-induced hepcidin [Ramey et al., 2009, PMID 19454495]. TFR2/HFE also up-regulate furin via MAPK/ERK, and furin maturation of BMPs feeds back onto hepcidin [Poli et al., 2010, PMID 20634490].
- Net output: Hepcidin protein is secreted, binds the iron exporter ferroportin (SLC40A1) on enterocytes and macrophages, and triggers its degradation, lowering dietary iron absorption and iron recycling [Kawabata 2018, PMID 29134618].
Interpretation: TFR2 is the "iron-replete" arm of the hepcidin thermostat — high transferrin saturation → stabilized TFR2 → ERK+SMAD signaling → high hepcidin → less iron entering plasma. This directly explains why loss of TFR2 causes iron overload.
Direct feature analysis of the UniProt record (Q9UP52, 801 aa) defines the architecture and provides bioinformatic evidence for the non-catalytic nature of TFR2:
- Topology: a type II single-pass membrane protein — short cytoplasmic N-terminus (residues 1–83, including a disordered region 16–45 and an endocytosis-signal motif at 23–26), a signal-anchor transmembrane helix (84–104), and a large extracellular C-terminal ectodomain (105–801) that carries the PA (protease-associated) domain, the peptidase-M28-like domain, and the TfR-like dimerization domain.
- N-glycosylation at N240, N339, N540, N754 — matching experimental mapping (N240/339/754 used; N540 not) [PMID 23556518].
- No catalytic machinery: the record annotates no "Active site" and no "Metal binding" (catalytic zinc) residues. Family M28 peptidases are normally zinc-dependent proteases; their absence in TFR2 is direct evidence that it is a non-catalytic "pseudo-peptidase."
Evolutionary interpretation: TFR2 arose from the same transferrin-receptor lineage as TFR1 (TFRC). The ancestral zinc-peptidase (M28/PA) fold has been repurposed as a ligand-binding receptor scaffold for holo-transferrin rather than to hydrolyze peptide bonds — a classic case of an enzyme fold evolving into a signaling/receptor module. This reconciles the UniProt "peptidase M28B" family assignment with the complete absence of any reported proteolytic activity in the experimental literature.
The TFR2 gene produces two main transcripts: full-length α-TFR2 and a shorter β-TFR2. The two partition function by tissue [Roetto et al., 2010, PMID 20179178]:
- α-TFR2 is the classical diferric-transferrin sensor that modulates hepcidin in the liver; hepatic α-TFR2 loss reproduces liver iron overload with inadequate hepcidin.
- β-TFR2, expressed in spleen, appears to control splenic iron efflux — β-specific knock-in mice have normal transferrin saturation/hepcidin but severe spleen iron accumulation with strikingly reduced splenic ferroportin (Fpn1), suggesting β-TFR2 influences Fpn1 transcription independently of the hepcidin axis.
This isoform split refines the "single receptor" picture: the hepatic hepcidin-sensing function is an α-TFR2 property, while extrahepatic/splenic iron handling involves β-TFR2.
We de-emphasize broad pleiotropic effects; the two axes above are the precise, mechanistically supported roles.
The signal TFR2 helps generate converges on the hepcidin–ferroportin axis: secreted hepcidin binds ferroportin (SLC40A1), the only cellular iron exporter, and induces its degradation, thereby reducing iron entry into plasma [Nai et al., 2025, PMID 40603805]. Because TFR2-driven hepcidin sits upstream of this rate-limiting step, TFR2 loss (low hepcidin) causes iron overload, while pharmacological restoration of hepcidin is being pursued therapeutically for hemochromatosis and β-thalassemia [PMID 40603805]. This positions TFR2 as the iron-replete sensor input to a druggable hormonal axis.
| Hypothesis | Verdict | Key evidence |
|---|---|---|
| TFR2 is a diferric-transferrin sensor that positively regulates hepcidin | Supported | PMID 41662592, 15319290, 29134618 |
| TFR2 signals via a surface HFE–TFR2–HJV complex through ERK + BMP/SMAD | Supported | PMID 22728873, 19454495, 20634490 |
| TFR2 has a distinct erythroid role via EPOR and intracellular iron routing | Supported | PMID 25499454, 32054685, 29969719 |
| TFR2 (as an M28-family member) is a catalytically active peptidase | Refuted / no evidence | No proteolytic activity reported; family M28B receptor branch is non-catalytic |
| TFR2 is a major route of transferrin-bound iron uptake (like TFR1) | Refuted | Only TFR1-expressing hepatocytes internalize holo-Tf; TFR2 contributes negligibly to iron uptake (PMID 41662592) |
Report compiled across Iterations 1–5. Evidence prioritizes primary hepatocyte/erythroid cell studies, targeted mutagenesis, knockout models, and direct UniProt feature analysis over high-throughput data. 26 papers reviewed; 8 findings recorded.