Ioav Cabantchik

Iron chelators are important tools in biochemical studies of iron metabolism and in the therapy of iron overload diseases. Their mode of action is comprised of entry into cells and scavenging intracellular metal, which includes complexation and egress of the complex. Iron is a metal which appears in the cells in various chemical forms and in different compartments. The form of the metal directly affected by the chelators is the most labile, low-molecular-weight type, which is present in the cytosol and is known as the "chelatable iron." This form is thought to be in dynamic equilibrium with several sequestered forms present in the cell, including the iron-responsive proteins. We recently introduced a fluorescent method for assessing the chelatable iron pool of cells, based on the quenching of the fluorescent calcein by metal ions (Breuer et al., J. Biol. Chem., 1995, 270, 24209-24215). In this work we adapted the method for dynamic assessment of chelator efficacy in scavenging iron from cells. In assay 1, red blood cells ghosts are used as a cell membrane model. The free-acid (impermeant) form of calcein is loaded into ghosts by encapsulation (lysis and resealing) and its fluorescence is quenched by addition of permeant iron(II). Chelators added to ghosts lead to iron removal from calcein and hence to recovery of fluorescence, commensurate with their permeation into ghosts and iron binding affinity. In assay 2, human K562 erythroleukemia cells are loaded with calcein via its permeating and cleavable acetoxymethyl form. A fraction of the intracellular calcein fluorescence is quenched in situ by endogenous cellular iron. The rate of dequenching which is obtained after addition of a chelator provides a measure for the scavenging of the intracellular metal, a process which, as in assay 1, depends on chelator permeation and binding affinity for iron. The two methods provide convenient means for assessing the efficacy of candidate chelator structures in depleting cell iron pools. They are also potentially applicable to chelation of other metals such as Co(II), Ni(II), and Cu(II) and to any cellular system or membrane vesicles.

The disulfonic acid stilbene derivative SITS reported to be covalently bonded to the membrane of the red blood cell, was found to be largely reversibly bound. Reversal of its specific inhibitory effect on anion permeability was attained by washing the cells with buffer containing albumin. The small fraction of covalently bonded SITS could be increased by prolonging the time of exposure of the cells or by multiple exposures. A series of other disulfonic stilbene derivatives was synthesized. All of them specifically inhibited anion permeability whether or not they are capable of forming covalent bonds. Their inhibitory effectiveness, however, varied over a 5,000-fold range, allowing certain conclusions to be made concerning the chemical architecture of the binding site. Certain of the compounds were almost entirely covalently bonded. One of them was labeled with125I and used to determine to which membrane proteins the compound is bound. Over 90% was found in a protein band on acrylamide gels of 95,000 mol wt. The most effective compound against sulfate permeability was equally effective against chloride permeability, producing a maximum inhibition of over 95%. The residual anion fluxes respond differently to pH and temperature than do the fluxes of unmodified cells.

The concept of non-transferrin bound iron (NTBI) was introduced 22 years ago by Hershko et al. (Brit. J. Haematol. 40 (1978) 255). It stemmed from a suspicion that, in iron overloaded patients, the large amounts of excess iron released into the circulation are likely to exceed the serum transferrin (Tf) iron-binding capacity (TIBC), leading to the appearance of various forms of iron not bound to Tf. In accordance with this assumption, NTBI was initially looked for and detected in patients with > or = 100% Tf-saturation. As techniques for its detection became more sophisticated and sensitive, NTBI was also found in conditions where Tf was not fully saturated, leading to a revision of the original view of NTBI as a simple spillover phenomenon. In this review, we will discuss some of the properties of NTBI, methods for its detection, its significance and potential value as an indicator for therapeutic regimens of iron chelation and supplementation.

Labile plasma iron (LPI) associated with iron supplementation has been implicated in complications found in dialysis patients. As LPI can potentially catalyse oxygen radical generation, we determined the presence of labile iron in the parenteral preparations and the frequency of occurrence of LPI in dialysis patients. The capacity to donate iron to apotransferrin (apo-) or to the chelator desferrioxamine (DFO) was measured with fluorescein-Tf (Fl-Tf) and Fl-DFO, respectively. Those probes undergo quenching upon binding to iron. Iron-catalysed generation of oxidant species was determined with dihydrorhodamine. Plasma nontransferrin-bound iron (NTBI), here termed LPI, was determined by mobilization of iron from low-affinity binding sites with oxalate, followed by its quantification with Fl-Tf in the presence of Ga(III). Normal individuals and most (80%) dialysis patients, analysed at least 1 week after iron supplementation showed no detectable (<0.2 microm) LPI. However, approximately 20% of the patients (n = 71) showed significant LPI levels (>0.2 microm), in some cases weeks after iron administration. LPI levels correlated best (r2 = 0.9) with Tf saturation. The iron preparations contained 2-6% low molecular weight and redox-active iron, most of which is chelated by Tf. Parenteral iron formulations contain a small but significant fraction of redox-active iron, most of which is scavenged by apo-Tf within <1 h. Therefore, oxidant stress associated with iron infusion is likely to be transient. The bulk of the polymeric iron is apparently inaccessible to apo-Tf. Although LPI might return to normal within 2 h of intravenous iron infusion, the long-term persistence of low-level LPI in up to 20% of end stage renal disease (ESRD) patients indicates that complete clearance of the intravenous iron may be more protracted than originally estimated.

The anion exchange system of human red blood cells is highly inhibited and specifically labeled by isothiocyano derivatives of benzene sulfonate (BS) or stilbene disulfonate (DS). To learn about the site of action of these irreversibly binding probes we studied the mechanism of inhibition of anion exchange by the reversibly binding analogs p-nitrobenzene sulfonic acid (pNBS) and 4,4'-dinitrostilbene-disulfonic acid (DNDS). In the absence of inhibitor, the self-exchange flux of sulfate (pH 7.4, 25 degrees C) at high substrate concentration displayed self-inhibitory properties, indicating the existence of two anion binding sites: one a high-affinity transport site and the other a low-affinity modifier site whose occupancy by anions results in a noncompetitive inhibition of transport. The maximal sulfate exchange flux per unit area was JA = (0.69 +/- 0.11) X 10(-10) moles . min-1 . cm-2 and the Michaelis-Menten constants were for the transport site KS = 41 +/- 14 mM and for the modifier site Ks' = 653 +/- 242 mM. The addition to cells of either pNBS at millimolar concentrations or DNDS at micromolar concentrations led to reversible inhibition of sulfate exchange (pH 7.4, 25 degrees C). The relationship between inhibitor concentration and fractional inhibition was linear over the full range of pNBS or DNDS concentrations (Hill coefficient n approximately equal to 1), indicating a single site of inhibition for the two probes. The kinetics of sulfate exchange in the presence of either inhibitor was compatible with that of competitive inhibition. Using various analytical techniques it was possible to determine that the sulfate transport site was the target for the action of the inhibitors. The inhibitory constants (Ki) for the transport sites were 0.45 +/- 0.10 microM for DNDS and 0.21 +/- 0.07 mM for pNBS. From the similarities between reversibly and irreversibly binding BS and DS inhibitors in structures, chemical properties, modus operandi, stoichiometry of interaction with inhibitory sites, and relative inhibitory potencies, we concluded that the anion transport sites are also the sites of inhibition and of labeling of covalent binding analogs of BS and DS.

The nucleoside transport of systems of hamster cells are susceptible to inhibition by S-6-substituted derivatives of mercaptonucleosides. The mechanism of interaction between the most potent inhibitor of the class, 6-nitrobenzyl mercaptoinosine (NBMI) and the uridine transport system of hamster fibroblasts is studied in the present work. A kinetic description of the interaction is presented. Uridine transport is inhibited in a partially competitive fashion, leaving a substantial free fraction of the transport (20–30%) virtually insensitive to increasing concentrations of inhibitor. One interpretation compatible with the kinetic and chemical properties of the system assumes that binding of the inhibitor to carriers occurs at sites different from the substrate binding sites (allosteric binding). Such a binding induces a conformational change in the carrier as manifested in a reduced affinity to the substrate and a susceptibility to inhibition by organomercurials. The alternative interpretation postulates two parallel transport systems which display distinctly different susceptibilities to NBMI and to organomercurials. Experiments performed with non-penetrating organomercurials show that the sulfhydryl groups related either to the alleged allosteric components or to additional carriers, are located superficially on the membrane. The binding of NBMI is reversible, the affinity is extremely high (Ki = 0.15 nMolar) and the rate of reaction could probably be diffusionally limited at low concentrations of reagent (activation energy 250 cal/mole, rate k = 1.3.108 min−1. Molar−1). The high affinity properties of the probes are used to determine the number of NBMI binding sites. A value of 47,000 and 77,000 sites/cell was obtained by two separate methods.The possibility that allosteric properties are present in carrier systems are discussed in terms of current concepts of modulation of transport functions in biological membranes.

We demonstrate in this work that HCO inf3sup−uptake in the marine macroalga Ulva sp. features functional resemblances to anion transport mediated by anion exchangers of mammalian cell membranes. The evidence is based on (i) competitive inhibition of photosynthesis by the classical red-blood-cell anion-exchange blockers 4,4′-dinitrostilbene-2,2′-disulfonate and 4-nitro-4′-isothiocyanostilbene-2,2′-disulfonate under conditions where HCO inf3sup−, but not CO2, was the inorganic carbon form taken up; (ii) inhibition of HCO inf3−uptake by pyridoxal phospate, indicating the involvement of lysine residues in the binding/translocation of HCO inf3sup−; and (iii) inhibition of HCO inf3sup−(but not of CO2) uptake by exofacial trypsin treatments, indicating the functional involvement of a plasmalemma protein. It is suggested that HCO inf3sup−uptake mediated by such a putative anion transporter can be a fundamental step in providing inorganic carbon for the CO2-concentrating system of marine marcoalgae in an environment where the HCO inf3sup−concentration is high, but the CO2 concentration and rates of uncatalyzed HCO inf3sup−dehydration are low.

Human intraerythrocytic malarial parasites (Plasmodium falciparum) induce permeability changes in the membrane of their host cells. The differential permeability of infected erythrocytes at various stages of parastie growth, in combination with density gradient centrifugation, was used to fractionate parasitized cells according to their developmental stage. By this method it was possible to obtain cell fractions consisting essentially of erythrocytes infected with the youngest parasite stage (i.e., rings). These preparations were used for the measurement of transport of various solutes. It is shown that permeabilization of host erythrocyte membrane appers as early as 6 h after parasite invasion of the erythrocyte and increases gradually with parasite maturation. Since the selectivity for several different solutes and the enthalpy of activation of transport remain unaltered with maturation-related increase of permeability, it is concluded that the number of transport agencies in the host cell membrane increases with parasite maturation. Evidence is presented to indicate the need for parasite protein synthesis as an essential factor for the generation of the new permeability pathways.

(3H)DIDS (4,4′-diisothiocyano-2,2′-ditritiostilbene-disulfonate) was used as a convalent label for membrane sites involved in anion permeability. The label binds to a small, superficially located population of sites, about 300,000 per cell, resulting in almost complete inhibition of anion exchange. The relationship of biding to inhibition is linear suggesting that binding renders each site nonfunctional. In the inhibitory range less than 1% of the label is associated with lipids but at higher concentrations of DIDS, the fraction may be as high as 4%. In ghosts, however, treatment with (3H)DIDS results in extensive labeling of lipids. In cells, a protein fraction that behavens on SDS acrylamide gels as thought its molecular weight is 95,000 daltons (95K) is predominatly labeled by (3H)DIDS. The only other labeled protein is the major sialoglycoprotein which contains less than, 5% of the total bound (3H)DIDS. Because of the linear relationship of binding to inhibition and the unique architecture of the site, it is suggested that the (3H)DIDS-binding site of the 95K protein is the substrate binding site of the anion transport system. The 95K protein is asymmetrically arranged in the membrane with the sites arranged on the outer face accessible to agent in the medium. In “leaky” ghost, only a few additional binding sites can be reached from the inside of the membrane in the 95K protein, in contrast to the extensive labeling of other membrane proteins in ghosts as compared to cells.

β-Thalassaemia represents a group of diseases, in which ineffective erythropoiesis is accompanied by iron overload. In a mouse model of β-thalassaemia, we observed that the liver expressed relatively low levels of hepcidin, which is a key factor in the regulation of iron absorption by the gut and of iron recycling by the reticuloendothelial system. It was hypothesised that, despite the overt iron overload, a putative plasma factor found in β-thalassaemia might suppress liver hepcidin expression. Sera from β-thalassaemia and haemochromatosis (C282Y mutation) patients were compared with those of healthy individuals regarding their capacity to induce changes the expression of key genes of iron metabolism in human HepG2 hepatoma cells. Sera from β-thalassaemia major patients induced a major decrease in hepcidin (HAMP) and lipocalin2 (oncogene 24p3) (LCN2) expression, as well as a moderate decrease in haemojuvelin (HFE2) expression, compared with sera from healthy individuals. A significant correlation was found between the degree of downregulation of HAMP and HFE2 induced by β-thalassaemia major sera (r = 0·852, P < 0·0009). Decreased HAMP expression was also found in HepG2 cells treated with sera from β-thalassaemia intermedia patients. In contrast, the majority of sera from hereditary haemochromatosis patients induced an increase in HAMP expression, which correlated with transferrin (Tf) saturation (r = 0·765, P < 0·0099). Our results suggest that, in β-thalassaemia, serum factors might override the potential effect of iron overload on HAMP expression, thereby providing an explanation for the failure to arrest excessive intestinal iron absorption in these patients.

The proteolytic enzymes, pronase, chymotrypsin and trypsin, release a small fraction of covalently bonded 4,4′-diisothiocyano-2,2′-ditritiostilbene disulfonate or (3H)DIDS, a specific inhibitor of anion permeability, from intact human red cells. The rate of release is parallel to the digestion of the sialoglycoprotein, indicating that the released (3H)DIDS was bound to that component. Most of the label is associated with a protein that behaves on SDS acrylamide gel electrophoresis as though its molecular weight was 95,000 Daltons (95K). Trypsin has no effect on this protein, but after pronase or chymotrypsin treatment of the cells, the label is found in three peaks of 95, 65 and 35K in proportions of 5, 85 and 10%. In parallel, the enzyme treatment results in a shift of most of the 95K protein to 65 and 35K. The digestion of the glycoprotein and splitting of the 95K protein can occur without any appreciable effects of the enzymes on anion permeability or on the inhibitory effects of DIDS treatment either before or after proteolytic attack. It is suggested that the sialoglycoprotein and the 35K segment of the 95K protein are not involved directly in anion permeation. The most likely location of the inhibitory site is in a portion of the 65K segment, exposed to the outside surface. Some additional observations are presented concerning the shielding effects of the negatively charged sialoglycoprotein and the arrangement of the 95K protein in the membrane.

A1 adenosine-receptor-antagonist drugs such as 8-cyclopentyl-1,3-dipropylxanthine (CPX) and xanthine amine congener (XAC) are found to activate the efflux of 36Cl- from CFPAC cells. These cells are a pancreatic adenocarcinoma cell line derived from a cystic fibrosis (CF) patient homozygous for the common mutation, deletion of Phe-508. The active concentrations for these compounds are in the low nanomolar range, consistent with action on A1 adenosine receptors. In addition, drug action can be blocked by exogenous agonists such as 2-chloroadenosine and also can be antagonized by removal of endogenous agonists by treatment with adenosine deaminase. Cells lacking the CF genotype and phenotype, such as HT-29 and T84 colon carcinoma cell lines, appear to be resistant to activation of chloride efflux by either drug. CFPAC cells transfected with the CF transmembrane regulator gene, CFTR, are also resistant to activation by CPX. We conclude that, since these antagonists are of relatively low toxicity and appear to act somewhat selectively, they might be considered as promising therapeutic candidates for CF.

... J. Inorg. Biochem. 47 (1992), pp. 183–195. Abstract | PDF (1085 K) | View Record in Scopus | Cited By in Scopus (73). 5. A. Jacobs. Blood 50 (1977), pp. 4331–4336. 6. W. Breuer, S. Epsztejnand ZI Cabantchik. J. Biol. Chem. 270 (1995), pp. 24209–24215. ...

A variety of biochemical, pharmacological, and toxicological properties have been attributed to labile forms of iron that are associated with cells or with biological fluids. Unlike the major fraction of bioiron which is protein bound, the labile bioiron is chelatable and therefore amenable for detection by metal-sensing devices that are coupled to chelation moieties. The present review deals with the detection of various labile forms of iron present in living cells and in fluids of biological interest, in health and disease. The main focus of the review is on the design and application of fluorescent probes as analytical tools for assessing labile iron and iron transport mechanisms and the efficiency of iron chelators in solution and in cellular systems.

Abstract: In thalassemia, iron overload is the joint outcome of excessive iron absorption and transfusional siderosis. While iron absorption is limited by a physiologic ceiling of about 3 mg/d, plasma iron turnover in thalassemia may be 10 to 15 times normal, caused by the wasteful, ineffective erythropoiesis of an enormously expanded erythroid marrow. This outpouring of catabolic iron exceeds the iron-binding capacity of transferrin and appears in plasma as non-transferrin-plasma iron (NTPI). The toxicity of NTPI is much higher than of transferrin-iron as judged by its ability to promote hydroxyl radical formation resulting in peroxidative damage to membrane lipids and proteins. In the heart, this results in impaired function of the mitochrondrial respiratory chain and abnormal energy metabolism manifested clinically in fatal hemosiderotic cardiomyopathy. Ascorbate increases the efficacy of iron chelators by expanding the intracellular chelatable iron pool, but, at suboptimal concentrations is a pro-oxidant, enhancing the catalytic effect of iron in free radical formation. NTPI is removed by i.v. DFO in a biphasic manner and reappears rapidly upon cessation of DFO, lending support to the continuous, rather than intermittent, use of chelators. Unlike DFO and other hexadentate chelators, bidentate chelators such as L1 may produce incomplete intermediate iron complexes at suboptimal drug concentrations.

Epithelial chloride channels can be blocked by various inhibitors, which show considerable differences in their molecular structure. In the present patch-clamp study, we compared different blockers of one type of epithelial Cl− channel with respect to their inhibitory potency. We applied the blockers to excised inside-out-or outside-out-oriented membrane patches of cultured HT29 colon carcinoma and respiratory epithelial cells (REC) containing the outwardly rectifying intermediate-conductance (ICOR) chloride channel. Four types of inhibitory compounds were tested: stilbene disulphonate derivatives, indanyloxyacetic acid, amidine, and arylaminobenzoates. The concentrations for half-maximal inhibition (IC50) for the different channel blockers were (μmol/l): 4-acetamido-4′-isothiocyanato-stilbene-2,2′-disulphonic acid 100; 4,4′-diisothiocyanato-stilbene-2,2′-disulphonic acid 80; indanyloxyacetic acid 9; 4,4′-dinitrostilbene-2, 2′-disulphonic acid 8; amidine 8 and 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) 0.9. All compounds, when applied to the cytosolic side of the channel, induced a flicker-type block of the ICOR Cl− channel at lower concentrations and a complete channel inhibition at higher concentrations. The inhibitory potency of NPPB was much higher when it was added to the external surface of the channel in outside-out-oriented membrane patches. At 1 μmol/l the inhibition was complete. All blocker effects were fully reversible. The probe with the highest affinity (NPPB) and a closely related compound 5-nitro-2-(3-phenylethylamino)-benzoate (NPFB) were used to construct macromolecular probes by linking these blockers to aminopolyethyleneglycol (PEG) or aminoethyl-O-dextran (5 kDa). These macromolecular NPPB and NPEB derivatives inhibited the ICOR Cl− channels only from the outside but had no effect on the cytosolic side. In the case of PEG-NPPB an IC50 of 30 nmol/l was determined in outside-out patches. The data indicate that the interaction site for arylaminobenzoates is accessible from the outer aspects of the Cl− channel facing the extracellular medium. Furthermore, these data show that the macromolecular probes of arylaminobenzoates have affinities to the Cl− channel very similar to those of the respective parent compounds.

Malaria parasites growing inside human erythrocytes differ from mammalian cells in their mode of acquisition of bioavailable iron and in their susceptibility to the antiproliferative action of iron chelators. We have assessed here three major properties associated with these phenomena: (a) the stage-dependent nature of parasite iron mobilization from the host and its integration into parasite proteins; (b) the differential permeability of the plasma membrane to iron chelators, and (c) the in situ generation of toxic chelator-metal complexes in the intracellular milieu of infected cells. We have used a combination of synthetic and natural iron chelators with similar iron-binding properties but markedly different capacities to permeate membranes. The profiles of action of these agents on the in vitro growth of Plasmodium falciparum were assessed in terms of inhibitory concentrations, speed of action, stage dependence and reversibility of effects. These profiles provided the basis for a working model of chelator action on parasitized cells. The model allowed us to predict major improvements in the antimalarial performance of iron chelators when used in appropriate combinations of slow-and fast-permeating substances. The synergistic actions found in vitro for various combinations of iron chelators are in accordance with the model and have implications for the design of therapeutic schemes.