Binding to the transferrin receptor is required for endocytosis of HFE and regulation of iron homeostasis.

HFE, the protein that is mutated in hereditary haemochromatosis, binds to the transferrin receptor (TfR). Here we show that wild-type HFE and TfR localize in endosomes and at the basolateral membrane of a polarized duodenal epithelial cell line, whereas the primary haemochromatosis HFE mutant, and another mutant with impaired TfR-binding ability accumulate in the ER/Golgi and at the basolateral membrane, respectively. Levels of the iron-storage protein ferritin are greatly reduced and those of TfR are slightly increased in cells expressing wild-type HFE, but not in cells expressing either mutant. Addition of an endosomal-targeting sequence derived from the human low-density lipoprotein receptor (LDLR) to the TfR-binding-impaired mutant restores its endosomal localization but not ferritin reduction or TfR elevation. Thus, binding to TfR is required for transport of HFE to endosomes and regulation of intracellular iron homeostasis, but not for basolateral surface expression of HFE.

Co-migration and internalization of transferrin and its receptor on K562 cells.

The incorporation of iron into human cells involves the binding of diferric transferrin to a specific cell surface receptor. We studied the process of endocytosis in K562, a human erythroid cell line, by using tetramethylrhodamine isothiocyanate-labeled transferrin (TRITC-transferrin) and fluorescein isothiocyanate-labeled Fab fragments of goat antireceptor IgG preparation (FITC-Fab-antitransferrin receptor antibody). Because the antireceptor antibody and transferrin bind to different sites on the transferrin receptor molecule it was possible to simultaneously and independently follow ligand and receptor. At 4 degrees C, the binding of TRITC-transferrin or FITC-Fab antitransferrin receptor antibody exhibited diffuse membrane fluorescence. At 20 degrees C, the binding of TRITC-transferrin was followed by the rapid formation of aggregates. However, the FITC-Fab antitransferrin receptor did not show similar aggregation at 20 degrees C unless transferrin was present. In the presence of transferrin, the FITC-Fab antitransferrin receptor antibody formed aggregates at the same sites and within the same time period as TRITC transferrin, indicating co-migration. Although the diffuse surface staining of either label was removed by proteolysis, the larger aggregates were not susceptible to enzyme degradation, indicating that they were intracellular. The internal location of the aggregates was also demonstrated using permeabilized cells that had been preincubated with transferrin and fixed with 4% paraformaldehyde. These cells showed aggregated receptor in the interior of the cell when reacted with fluorescein-labeled antibody to the receptor. This indicated that the transferrin and the transferrin receptor co-internalize and migrate to the same structures within the cell.

Cytoskeletal protein ABP-280 directs the intracellular trafficking of furin and modulates proprotein processing in the endocytic pathway.

Furin catalyzes the proteolytic maturation of many proproteins within the trans-Golgi network (TGN)/endosomal system. Furin's cytosolic domain (cd) directs both the compartmentalization to and transit between its manifold processing compartments (i.e., TGN/biosynthetic pathway, cell surface, and endosomes). Here we report the identification of the first furin cd sorting protein, ABP-280 (nonmuscle filamin), an actin gelation protein. The furin cd was used as bait in a yeast two-hybrid screen to identify ABP-280 as a furin-binding protein. Binding analyses in vitro and coimmunoprecipitation studies in vivo showed that furin and ABP-280 interact directly and that ABP-280 tethers furin molecules to the cell surface. Quantitative analysis of both ABP-280-deficient and genetically replete cells showed that ABP-280 modulates the rate of internalization of furin but not of the transferrin receptor, a cycling receptor. However, although ABP-280 directs the rate of furin internalization, the efficiency of sorting of the endoprotease from the cell surface to early endosomes is independent of expression of ABP-280. By contrast, efficient sorting of furin from early endosomes to the TGN requires expression of ABP-280. In addition, ABP-280 is also required for the correct localization of late endosomes (dextran bead uptake) and lysosomes (LAMP-1 staining), demonstrating a pleiotropic role for this actin binding protein in the organization of cellular compartments and directing protein traffic. Finally, and consistent with the trafficking studies on furin, we showed that ABP-280 modulates the processing of furin substrates in the endocytic but not the biosynthetic pathways. The novel roles of ABP-280 and the cytoskeleton in the sorting of furin in the TGN/ endosomal system and the formation of proprotein processing compartments are discussed.

Interactions of the ectodomain of HFE with the transferrin receptor are critical for iron homeostasis in cells.

Expression of wild type HFE reduces the ferritin levels of cells in culture. In this report we demonstrate that the predominant hereditary hemochromatosis mutation, C282Y(2) HFE, does not reduce ferritin expression. However, the second mutation, H63D HFE, reduces ferritin expression to a level indistinguishable from cells expressing wild type HFE. Further, two HFE cytoplasmic domain mutations engineered to disrupt potential signal transduction, S335M and Y342C, were functionally indistinguishable from wild type HFE in this assay, as was soluble HFE. These results implicate a role for the interaction of HFE with the transferrin receptor in lowering cellular ferritin levels.

Pumping iron: the strange partnership of the hemochromatosis protein, a class I MHC homolog, with the transferrin receptor.

People suffering from hereditary hemochromatosis (HH) can not regulate the uptake of iron properly and gradually accumulate iron in their body over their lifetime. The protein involved in HH, HFE, has been recently identified as a class I major histocompatibility complex (MHC) homolog. The wild-type HFE associates and co-traffics with the transferrin receptor (TfR). The mutation responsible for 83% of HH (C260Y) results in the failure of HFE to form a critical disulfide bond, bind beta2 microglobulin, bind TfR, and traffic to the cell surface. In non-polarized cells, the partnership of HFE and TfR results in decreased iron uptake into cells. The mechanism whereby a class I MHC homolog modifies the function of a membrane receptor and how this dynamic complex of molecules regulates iron transport across intestinal epithelial cells is the subject of this review.

Structure and function of aspartate transcarbamoylase studied using chymotrypsin as a probe.

Aspartate transcarbamoylase from Escherichia coli is composed of six catalytic (c) and six regulatory (r) polypeptides. We have studied the structure and function of this enzyme using chymotrypsin as a probe. The protease inactivates the isolated catalytic subunit (c3) but has not effects on the native enzyme (c6r6). Under identical conditions, the c3r6 complex is inactivated at a much slower rate than c3. The presence of the substrate analogue succinate together with carbamoyl phosphate reduces substantially the rate of inactivation. Extended exposure to chymotrypsin converts the catalytic subunit into a partially active derivative with a fourfold higher Michaelis constant. This derivative is indistinguishable from the unmodified catalytic subnit in gell electrophoresis under nondenaturing conditions. However, in the presence of sodium dodecyl sulfate, the major fragment in the electropherogram is smaller than that of the intact catalytic polypeptide. The results could be explained by postulating the presence of a chymotrypsin-sensitive peptide bond at or near the active site. Since X-ray crystallographic studies have indicated that the active sites are located in a central cavity, the resistance of the native enzyme towards inactivation may be due to the inability of chymotrypsin to enter this cavity.

Aspartate transcarbamoylase: loss of homotropic but not heterotropic interactions upon modification of the catalytic subunit with a bifunctional reagent.

The role of conformational changes in the allosteric mechanism of aspartate transcarbamoylase from Escherichia coli was studied by reacting the isolated catalytic subunit with the bifunctional reagent tartryl diazide. Two derivatives differing moderately in substrate affinity were obtained depending on whether the reaction was conducted in the presence or absence of the substrate analogue succinate and carbamoyl phosphate. The modification was not accompanied by aggregation or dissociation. The modified catalytic subunits retained the ability to reassociate with unmodified regulatory subunits and produced hybrids similar in size to the native enzyme. These hybrids were appreciably sensitive to the allosteric effectors ATP and CTP but unlike native enzyme showed no cooperativity in substrate binding. The Michaelis constants of these hybrids for aspartate were intermediate between that of the isolated catalytic subunit and that of the relaxed state. Activation by ATP was caused by a reduction in Km to the value characteristic of the relaxed state whereas CTP inhibited by lowering the Vmax. The properties of the hybrids are strikingly similar to the modified enzyme obtained by Kerbiriou and Hervé from cells grown in the presence of 2-thiouracil. However, the crucial modifications are found in the regulatory subunits of the enzyme studied by these authors whereas they are located in the catalytic subunits of the hybrids reported here. Our results suggest that interactions between the catalytic and regulatory subunits have considerable effects on the state of the substrate binding sites in the native enzyme.

Characterization of the transferrin receptor in tunicamycin-treated A431 cells.

The transferrin receptor undergoes extensive co- and post-translational modifications during its biosynthesis. In this study, the functional and structural properties of the transferrin receptor from tunicamycin-treated A431 cells were examined. Incubation of A431 cells with this inhibitor of asparagine-linked glycosylation results in a shift of the apparent molecular weight of the transferrin receptor from 94,000 to 79,000. The electrophoretic mobility of the receptor from treated cells is that of a monomer under nonreducing conditions, whereas the transferrin receptor in untreated cells has the mobility of a dimer under identical conditions. This result indicates a lack of disulfide bond formation between subunits of the receptor from tunicamycin-treated cells. In solution no dimers can be detected with cross-linking studies. This unglycosylated receptor does not appear to stably bind transferrin as demonstrated by a lack of isolation of this form of the receptor with transferrin-linked Sepharose. It is not transported to the surface of A431 cells.

A mutated transferrin receptor lacking asparagine-linked glycosylation sites shows reduced functionality and an association with binding immunoglobulin protein.

The function of the transferrin receptor is to transport iron-bound transferrin into the cell. In order to function properly, this dimeric glycoprotein must be expressed on the cell surface and be able to bind transferrin. Site-directed mutagenesis was performed to abolish the three asparagine-linked glycosylation consensus sequences of the human transferrin receptor. The DNA encoding the mutated transferrin receptor was stably transfected into mouse fibroblasts. This form of the human transferrin receptor shows reduced transferrin binding, reduced intersubunit bond formation, and reduced cell surface expression, indicating that the transferrin receptor which lacks asparagine-linked glycosylation is not fully functional. In addition, the mutated form of the receptor is not processed as quickly. It shows an association with an endoplasmic reticular chaperone protein, binding immunoglobulin protein, leading to the hypothesis that the mutated transferrin receptor experiences increased retention in the endoplasmic reticulum.

Iron homeostasis: new tales from the crypt.

The enterocyte is a highly specialized cell of the duodenal epithelium that coordinates iron uptake and transport into the body. Until recently, the molecular mechanisms underlying iron absorption and iron homeostasis have remained a mystery. This review focuses on the proteins and regulatory mechanisms known to be present in the enterocyte precursor cell and in the mature enterocyte. The recent cloning of a basolateral iron transporter and investigations into its regulation provide new insights into possible mechanisms for iron transport and homeostasis. The roles of proteins such as iron regulatory proteins, the hereditary hemochromatosis protein (HFE)-transferrin receptor complex, and hephaestin in regulating this transporter and in regulating iron transport across the intestinal epithelium are discussed. A speculative, but testable, model for the maintenance of iron homeostasis, which incorporates the changes in the iron-related proteins associated with the life cycle of the enterocyte as it journeys from the crypt to the tip of the villous is proposed.

Cleavage of the transferrin receptor is influenced by the composition of the O-linked carbohydrate at position 104.

A soluble form of the human transferrin receptor (TfR) resulting from proteolytic cleavage at Arg 100 has been measured in human blood. In tissue culture cells elimination of the O-linked carbohydrate at Thr 104, four amino acids from the cleavage site, results in enhanced cleavage of the TfR (Rutledge et al., 1994, Blood, 83:580-586). In the present set of studies, the influence of amino acid substitution and the composition of the oligosaccharide at amino acid 104 on the cleavage of the TfR was examined. Site-directed mutagenesis was used to generate six different amino acids at position 104 which varied in size and charge. Measurement of the soluble TfR in the conditioned medium of the transfected cells of each mutant TfR showed that the large and charged side chains inhibited TfR cleavage the most. Otherwise the properties of the mutant TfRs were indistinguishable from the wild-type TfR in that the affinity of transferrin for these receptors, the extent of disulfide bond formation of the TfRs, and the proportion of TfRs at the cell surface were similar to that of the wild-type TfR. Removal of the sialic acid component of the carbohydrate from wild-type TfR by treatment of live cells with neuraminidase enhances TfR cleavage. Expression of wild-type TfR in CHO IdlD cells (a glycosylation defective cell line) also shows enhanced cleavage under conditions that produce truncated or no O-linked carbohydrates. Treatment of IdlD cells with neuraminidase reveals that the sialic acid of the O-linked carbohydrate protects against TfR cleavage, whereas the core sugars Gal-NAc and Gal do not protect as much. These results show that the terminal charged sialic acid residues are important for protection from proteolytic cleavage and suggest that cleavage could be regulated in the cell by removal of all or part of the carbohydrate.

Stabilization of the relaxed state of aspartate transcarbamoylase by modification with a bifunctional reagent.

Native aspartate transcarbamoylase from Escherichia coli was modified with the bifunctional reagent tartaryl diazide in the presence of the substrate carbamoyl phosphate and the substrate analog succinate. The product had the same sedimentation coefficient as the native enzyme but showed a marked increase in affinity for the substrate aspartate with a hyperbolic saturation curve. The Michaelis constant for aspartate (7.4 mM) is similar to that estimated for the relaxed state of the enzyme. The high substrate affinity was not produced if modification was conducted in the absence of substrate analogs or with a monofunctional reagent. The modified enzyme was also desensitized towards the allosteric effectors ATP and CTP. It appears to represent a stabilized relaxed state whose conversion to the taut state is presumably prevented by cross-linking.

Hybrid aspartate transcarbamoylase containing cross-linked subunits.

The role of conformational changes and subunit interactions in the allosteric mechanism of aspartate transcarbamoylase was evaluated by studying hybrid enzyme molecules containing cross-linked subunits. Native enzyme was cross-linked with tartryl diazide in the presence and absence of substrate analogues. The two types of modified enzyme derivatives were each dissociated into catalytic (c3) and regulatory (r2) subunits. Hybrids were constructed with modified catalytic subunits and unmodified regulatory subunits of vice versa. Subunits from different derivatives also formed hybrids. All hybrids containing cross-linked catalytic subunits showed hyperbolic substrate saturation curves while cross-linked in the regulatory subunit alone did not abolish cooperativity. The type of cross-linked in the catalytic subunit had a decisive influence on the substrate affinity of the hybrid as well as its response to the allosteric effectors ATP and CTP. However many effects were also dependent on the presence of regulatory subunits. The results implicate a substantial conformational change in the catalytic subunit upon substrate binding and suggest an important role for the c-r interaction in the allosteric mechanism.

Characterization of the transferrin-binding protein in a human trophoblast.

The physical properties and binding characteristics of the solubilized transferrin-binding protein from BeWo cells, a human choriocarcinoma cell line, were investigated. The binding protein was isolated from 125I-labelled membranes by solubilization followed by immunoprecipitation with anti-human transferrin in the presence of saturating human transferrin. Gel filtration on acrylamide agarose (AcA-22) at 21 degrees C in the absence of transferrin indicates that the transferrin-binding protein has a Stokes' radius of 4.6 nm. In the presence of transferrin, the Stokes' radius of the transferrin-binding BeWo protein increases to 6.3 nm. Parallel sucrose density centrifugation studies indicate that the BeWo protein has a sedimentation coefficient of 9.4 S in the absence of transferrin and 10.9 S in the presence of transferrin. Relative molecular mass calculations from sedimentation studies in H2O and D2O, using the method of Sadler et al (1979), indicate a relative molecular mass of 204,000 for the solubilized receptor and 354,000 for the receptor in the presence of transferrin.

Similarities between the transferrin receptor proteins on human reticulocytes and human placentae.

The transferrin receptor of the human reticulocyte was isolated by two different immunoaffinity procedures. These included indirect immunoprecipitation with a transferrin/anti-transferrin complex and direct immunoprecipitation with antiserum to purified transferrin receptor from placentae. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the receptor isolated from reticulocytes reveals a polypeptide at Mr = 94,000 identical in molecular weight with that of the placenta. A radioimmunoassay using purified 125I-labeled transferrin receptor from placentae and antiserum to transferrin receptor fails to distinguish any immunological differences between the reticulocyte and placental forms of the protein. In addition, proteolytic digests of both of these polypeptides with Staphylococcus aureus protease show identical proteolytic patterns, indicating similar sequences.

Physical characterization of the transferrin receptor in human placentae.

The physical properties and binding characteristics of the solubilized transferrin receptor isolated from the placental brush-border membrane of a human trophoblast cell were investigated. The receptor protein was isolated from solubilized 125I-labeled membranes by immunoprecipitation with anti-human transferrin in the presence of saturating amounts of human transferrin. Gel filtration on acrylamide agarose (AcA-22) at 23 degrees C in the absence of transferrin indicates the transferrin receptor has a Stokes radius of 4.6 nm. In the presence of transferrin, the Stokes radius of the receptor shifts to 6.3 nm. Sucrose density centrifugation studies indicate that it has a sedimentation coefficient of 9.8 S in the absence of transferrin and 11.2 S in the presence of transferrin. The molecular weight for the transferrin free receptor is calculated to be 213,000. Upon incubation with transferrin, it increases to 364,000. This is consistent with the idea that the active form of the solubilized receptor is a dimer and the dimer is in turn capable of binding two transferrin molecules.

A region of the C-terminal portion of the human transferrin receptor contains an asparagine-linked glycosylation site critical for receptor structure and function.

The transferrin receptor is a cell surface protein and is responsible for the uptake of iron into many eukaryotic cells. In its mature form, the receptor possesses three asparagine-linked oligosaccharides. The effect of asparagine-linked glycosylation on the processing and cell surface localization of the human transferrin receptor is examined here by site-directed mutagenesis. Each of the extracellular consensus sequences (Asn-X-Ser/Thr) for asparagine-linked glycosylation was mutated individually and in all possible combinations. The constructs were transfected stably into NIH-3T3 cells and a Chinese hamster ovary cell line lacking endogenous transferrin receptors. Of the seven possible combinations of glycosylation sites, single mutations eliminating glycosylation at either Asn251 or Asn317 do not affect the processing and surface localization of the receptor. Eliminating both of these sites together has a small effect on the behavior of the receptor. However, mutation of the C-terminal glycosylation site (Asn727) has the most profound negative effect on the appearance of the receptor at the cell surface. The mutants lacking glycosylation at Asn727 appear to be retained in the endoplasmic reticulum as an increased association with binding immunoglobulin protein (BiP) is observed. Addition of a new glycosylation site in the C-terminal region of the unglycosylated mutated transferrin receptor restores the cell surface localization and the transferrin binding of the transferrin receptor, indicating that glycosylation in this region is critical for the correct transport of this receptor to the cell surface.

A rapid redistribution of the transferrin receptor to the cell surface of HL-60 cells and K562 cells upon treatment with dimethyl sulfoxide due to slowing of endocytosis.

Treatment of two human leukemia cell lines with 1.25% dimethyl sulfoxide at 37 degrees C results in a rapid increase in the number of transferrin receptors on the cell surface detected by fluorescein-labeled anti-transferrin receptor antibodies. Both HL-60 cells, a human myeloid cell line, and K562 cells, a human erythroid-myeloid cell line, showed a 25-65% increase in cell surface transferrin binding in parallel experiments. Scatchard plot analysis of the data indicates that the number of receptors increases while the affinity of transferrin for the receptor remains the same. This rapid increase in the number of receptors at the cell surface appears to be due to a slowing of endocytosis rather than an increase in externalization of the receptor.

Generation of the soluble transferrin receptor requires cycling through an endosomal compartment.

The transmembrane protein, transferrin receptor (TfR), is found in a soluble form in human serum and in the medium of cell lines grown in tissue culture. The soluble form is generated by proteolytic cleavage between Arg-100 and Leu-101. We used two mutant human TfRs expressed in Chinese hamster ovary (CHO) cells lacking endogenous transferrin receptor to characterize the protease that cleaves the TfR and determine its location in the cell. The T104D mutant TfR lacks the O-linked carbohydrate at position 104, and is more susceptible to proteolytic cleavage at Arg-100 than the wildtype human TfR in these cells. We find that the protease is not a component of the serum in the growth medium, and it is not secreted by the cells. Cleavage does not occur during biosynthesis of the TfR, and occurs after the TfR has reached the cell surface. Expression of the T104D TfR in a temperature-sensitive acidification defective CHO cell line, G.7.1, shows that cleavage of the TfR is not dependent on acidification of endosomes. The C20A23 TfR is an endocytosis deficient mutant lacking an internalization signal. This mutant TfR, which is mainly localized to the cell surface, is cleaved less efficiently than the wild-type TfR, indicating that the protease is localized to an intracellular compartment.