3-Hydroxy-(4H)-benzopyran-4-ones as potential iron chelating agents in vivo.

Increasing evidence suggests that iron plays an important role in tissue damage both during chronic iron overload diseases (i.e., hemochromatosis) and when, in the absence of actual tissue iron overload, iron is delocalised from specific carriers or intracellular sites (inflammation, neurodegenerative diseases, post-ischaemic reperfusion, etc.). In order to be used for therapeutical purposes in vivo, a reliable iron chelator should be capable of preventing the undesired effects that follow the electrochemical activation of iron (see below). Bearing in mind the molecular structure of some flavonols that are able to chelate iron, we synthesised a new oral iron-chelator, 2-methyl-3-hydroxy-4H-benzopyran-4-one (MCOH). We demonstrate that MCOH chelates iron in a 2:1 ratio showing a stability constant of approximately 10(10). MCOH is able to cross cell membranes (erythrocytes, ascite tumour cells) in both directions. Following intraperitoneal administration to rats, it is quickly taken up by the liver and excreted in the urine within 24h. A similar behaviour has been documented after oral administration. We propose that MCOH may represent the prototype of a new class of iron chelating agents to be developed for iron-removal therapy in vivo with the goal of preventing tissue damage caused by the iron redox cycle.

Iron release and oxidant damage in human myoblasts by divicine.

Divicine is an aglycone derived from vicine, a glucosidic compound contained in fava beans (Vicia faba major or broad beans). In this study, we investigated the effect of divicine on cultured human myoblasts from normal subjects, in order to see if the drug may induce signs of oxidant stress in these cells. Myoblasts incubated 24 hours in the presence of 1 mM divicine, showed an increase of carbonyl groups and 4-hydroxynonenal (4-HNE) bound to cell proteins, as well as a significant release of iron and lactate dehydrogenase in the culture medium. Desferrioxamine (DFO), an iron chelator, significantly prevented protein oxidation and formation 4-HNE adducts. Our results can be interpreted as indicating that divicine autooxidizes both at extracellular level and into myoblasts thus inducing the release of free iron, which initiates oxidation of cellular proteins and lipids. DFO protects the cells by subtracting the free iron both at intracellular and extracellular level.

Protection of erythrocytes against oxidative damage and autologous immunoglobulin G (IgG) binding by iron chelator fluor-benzoil-pyridoxal hydrazone.

Iron is released in a free desferrioxamine-chelatable form when erythrocytes are challenged by an oxidative stress. The release of iron is believed to play an important role in inducing destructive damage (lipid peroxidation and hemolysis) or in producing membrane protein oxidation and generation of senescent cell antigens (SCA). In this report, we further tested the hypothesis that intracellular chelation of iron released under conditions of oxidative stress prevents erythrocyte damage or SCA formation. Fluor-benzoil-pyridoxal hydrazone (FBPH), an iron-chelating molecule of the family of aromatic hydrazones, was prepared by synthesis and used for the above purpose after the capacity of the product to enter cells had been ascertained. GSH-depleted mouse erythrocytes were incubated with the oxidant drug phenylhydrazine in order to produce iron release, lipid peroxidation, and hemolysis. FBPH at a concentration of 200 microM prevented lipid peroxidation and hemolysis in spite of equal values of iron release. FBPH was active even at a lower concentration (100 microM) when the erythrocytes were preincubated with it for 15 min. No preventive effect was seen when FBPH saturated with iron was used. Prolonged aerobic incubation (60 hr) of erythrocytes produced iron release and formation of SCA as determined by autologous immunoglobulin G (IgG) binding. The IgG binding was detected by using an anti-IgG antibody labeled with fluorescein and by examining the cells for fluorescence by confocal microscopy. FBPH prevented SCA formation in a dose-related manner. These results lend further support to the hypothesis that iron release is a key factor in erythrocyte ageing.

Hemolytic drugs aniline and dapsone induce iron release in erythrocytes and increase the free iron pool in spleen and liver.

Incubation of rat erythrocytes with the hydroxylated metabolites of aniline and dapsone (4-4'-diaminodiphenylsulfone), phenylhydroxylamine and dapsone hydroxylamine, respectively, induced marked release of iron and methemoglobin formation. On the contrary, no release of iron nor methemoglobin formation was seen when the erythrocytes were incubated with the parent compounds (aniline and dapsone). The acute intoxication of rats with aniline or dapsone induced a marked increase in the erythrocyte content of free iron and methemoglobin, indicating that the xenobiotics are effective only after biotransformation to toxic metabolites in vivo. Prolonged administration of aniline or dapsone to rats produced continuous release of iron from erythrocytes. Marked iron overload was seen in the spleen and in the liver Kupffer cells, as detected histochemically. The spleen weight in these subchronically treated animals was significantly increased. The free iron pool was markedly increased in the spleen and to a lower extent in the liver. The possible relationships between iron release in erythrocytes, oxidative damage seen in senescent cells, hemolysis, overwhelmed capacity of spleen and liver to keep iron in storage forms and subsequent increase in low molecular weight, catalitically active iron is discussed.

Iron release in erythrocytes from patients with beta-thalassemia.

Our previous studies have shown that iron is released in a free (desferrioxamine-chelatable) form when erythrocytes undergo oxidative stress (incubation with oxidizing agents or aerobic incubation in buffer for 24-60 h (a model of rapid in vitro ageing)). The release is accompanied by oxidative alterations of membrane proteins as well as by the appearance of senescent antigen, a signal for termination of old erythrocytes. In hemolytic anemias by hereditary hemoglobin alterations an accelerated removal of erythrocytes occurs. An increased susceptibility to oxidative damage has been reported in beta-thalassemic erythrocytes. Therefore we have investigated whether an increased iron level and an increased susceptibility to iron release could be observed in the erythrocytes from patients with beta-thalassemia. Erythrocytes from subjects with thalassemia intermedia showed an extremely higher content (0 time value) of free iron and methemoglobin as compared to controls. An increase, although non-statistically-significant, was seen in erythrocytes from subjects with thalassemia major. Upon aerobic incubation for 24 h the release of iron in beta-thalassemic erythrocytes was by far greater than in controls, with the exception of thalassemia minor. When the individual values for free iron content (0 time) seen in thalassemia major and intermedia were plotted against the corresponding values for HbF, a positive correlation (P < 0.001) was observed. Also, a positive correlation (P < 0.01) was seen between the values for free iron release (24 h incubation) and the values for HbF. These results suggest that the presence of HbF is a condition favourable to iron release. Since in beta-thalassemia the persistance of HbF is related to the lack or deficiency of beta chains and therefore to the excess of alpha chains, the observed correlation between free iron and HbF, is consistent with the hypothesis by others that excess of alpha chains represents a prooxidant factor.

Circulating pro- and antioxidant factors in iron and porphyrin metabolism disorders.

BACKGROUND: Porphyria cutanea tarda and haemochromatosis are taken to be spontaneous human models of oxidative cellular damage, with an increased risk of fibrosis and cancer evolution. AIM: To define the relative pro-oxidant roles of porphyrin and iron, in their different molecular forms, and their effects on antioxidant biological systems. PATIENTS: A group of 17 patients with porphyria cutanea tarda and a group of 14 patients with primary and secondary haemochromatosis, were compared with 21 healthy controls. METHODS: Plasma retinol, tocopherol, alpha- and beta-carotene, ascorbic acid, glutathione, malonyldialdehyde and red blood cell free iron were determined using high performance liquid chromatography. RESULTS: Only a modest increase in iron stores was demonstrated in the porphyria cutanea tarda group; in the haemochromatosis patients ferritin levels were almost seven times higher. By contrast, there was a sharp and virtually identical increase in red blood cell free iron and malonyldialdehyde in both the patient groups. A significant reduction was observed in retinol, alpha-, beta-carotene and red blood cell glutathione levels being more marked in porphyria cutanea tarda than in haemochromatosis patients. CONCLUSIONS: The study confirms the strong pro-oxidant effects of porphyrins in vivo, through an induction of the free toxic iron form, even though the total iron pool is not greatly expanded. The additional free-iron and porphyrin oxidant effects are documented both in red blood cell and plasma in the porphyria cutanea tarda group. It confirmed that aging exerts a negative influence in terms of pro- and antioxidant balance in all cases, but particularly in the haemochromatosis group.

Effect of deferoxamine and allopurinol on non-protein-bound iron concentrations in plasma and cortical brain tissue of newborn lambs following hypoxia-ischemia.

Reduction of non-protein-bound iron (NPBI) using iron chelators may attenuate hypoxia-ischemia-induced reperfusion injury of the brain. This study investigated whether administration of low-dose deferoxamine and allopurinol, both having NPBI-chelating properties, reduced hypoxia-ischemia-induced NPBI formation in plasma effluent from the brain and in cerebral cortical tissue. Twenty-one newborn lambs underwent severe hypoxia-ischemia. Upon reperfusion and reoxygenation the lambs received either a placebo (n = 7), or deferoxamine 2.5 mg/kg (n = 7) or allopurinol 20 mg/kg (n = 7). The post-hypoxic-ischemic NPBI levels in plasma were significantly lower after deferoxamine but not after allopurinol as compared to placebo-treated lambs. Cortical NPBI levels in both deferoxamine and allopurinol-treated lambs were significantly lower than NPBI levels in placebo-treated lambs. We conclude that deferoxamine effectively lowers NPBI in plasma effluent from the brain, and that both, deferoxamine and allopurinol, lower NPBI in cortical brain tissue.

Protection against oxidative damage of erythrocyte membrane by the flavonoid quercetin and its relation to iron chelating activity.

Incubation of glutathione (GSH) depleted mouse erythrocytes with the oxidants phenylhydrazine, acrolein, divicine and isouramil resulted in the release of free iron and in lipid peroxidation and hemolysis. The addition of the flavonoid quercetin, which chelates iron and penetrates erythrocytes, resulted in remarkable protection against lipid peroxidation and hemolysis. The protection seems to be due to intracellular chelation of iron, since a semi-stoichiometric ratio between released iron and the amount of quercetin necessary to prevent lipid peroxidation and hemolysis was found. Incubation of GSH depleted human erythrocytes with divicine and isouramil did not induce lipid peroxidation and hemolysis in spite of a substantial release of iron. However, divicine and isouramil produced alterations of membrane proteins, such as spectrin and band 3, as well as formation of senescent cell antigen. The addition of quercetin prevented these alterations.

Iron mobilization from crocidolite as enhancer of collagen content in rat lung fibroblasts.

Asbestos exposure causes pulmonary fibrosis by mechanisms that remain uncertain. There is increasing evidence that iron from asbestos is responsible for many of its effects. In this paper, we investigated the effect of iron mobilized from crocidolite asbestos on collagen content in rat lung fibroblast cultures under serum-free conditions. Crocidolite (2, 4, 6 microg/cm2 well) increased collagen content in a dose-dependent manner (+42 +/- 8, +92 +/- 10, and +129 +/- 13% vs controls). This effect was specific for collagen, since it did not alter total protein content and was inhibited by the iron chelator deferoxamine (DFO). Preincubation of crocidolite with citrate (1 mM) for 48 hr resulted in iron mobilization (51 microM) and increased collagen production (>3-fold) in treated cells. These effects occurred without the intervention of serum factors. The absence of cell damage, proliferation or lipid peroxidation leads to the supposition that iron from crocidolite per se may act as a profibrogenic agent. Although the in vivo participation of other cells and factors cannot be excluded, we conclude that iron released from crocidolite plays a role in collagen increase occurring during asbestosis.

Release of free, redox-active iron in the liver and DNA oxidative damage following phenylhydrazine intoxication.

Following the subchronic intoxication of rats with phenylhydrazine, resulting in marked anemia, reticulocytosis, methemoglobinemia and increased hemocatheresis, the hepatic content of total iron was increased, as was hepatic ferritin and its saturation by iron. A striking increase (approximately 7-fold) was also observed in free iron which appeared to be redox-active. The increase in liver free iron involved the hepatocellular component of the liver. Since DNA is one of the cellular targets of redox active iron, liver DNA from phenylhydrazine-treated rats was analyzed by electrophoresis and found to be markedly fragmented. Experiments with isolated hepatocytes in culture or in suspension challenged with phenylhydrazine or Fe-nitrilotriacetate strongly suggested that the DNA damage was due to reactive iron rather than to the hepatic metabolism of phenylhydrazine. The levels of 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodGuo), a specific marker of oxidative DNA damage, were significantly higher in phenylhydrazine-treated rats as compared to untreated controls. The prolongation of phenylhydrazine treatment over a period of 6 weeks resulted in a persistent damage to DNA and in phenotypic changes such as an increase in hepatocyte gamma-glutamyl transpeptidase (gamma-GT, EC 2.3.2.2) activity. Possible relationships between iron overload, iron release, DNA damage and tumor initiation are discussed.

Relationship between free iron level and rat liver mitochondrial dysfunction in experimental dietary iron overload.

The concentration of total iron in the hepatic tissue and mitochondria from rats fed a 2.5% carbonyl iron supplemented diet progressively increased up to 40 days, then reached nearly a steady-state. By contrast the level of free iron (desferrioxamine-chelatable) exhibited a transient but significant increase at 40 days of treatment, only in this period of treatment the induction of lipid peroxidation and the resulting mitochondrial abnormalities in calcium transport was observed too. The enhancement of the energy dissipating mitochondrial calcium cycling was found to be associated with a significant decrease of endogenous mitochondrial ATP content. As to the pathophysiological mechanism for hepatocellular injury in iron overload, these results indicated that the transit pool of free iron may play a critical role in initiating organelle dysfunctions, at least in this experimental model of iron overload.

Iron release, membrane protein oxidation and erythrocyte ageing.

The aerobic incubation of erythrocytes in phosphate buffer for 24-60 h (a model of rapid in vitro ageing) induced progressive iron release and methemoglobin formation. Membrane proteins showed electrophoretic alterations and increase in carbonyl groups (as documented by IR spectroscopy). None of these phenomena were seen when the erythrocytes were incubated under anaerobic conditions. The membranes from aerobically incubated cells bound a much higher amount of autologous IgG than those from anaerobically incubated ones, suggesting that the aerobic incubation gives rise to the senescent antigen. The addition of ferrozine during the aerobic incubation prevented both the IgG binding and the protein alterations seen in the IR spectra, suggesting an intracellular chelation of the released iron by ferrozine.

Iron release, lipid peroxidation, and morphological alterations of erythrocytes exposed to acrolein and phenylhydrazine.

Iron is released in a free [desferrioxamine (DFO)-chelatable] form in mouse erythrocytes incubated with the oxidizing agents acrolein and phenylhydrazine or in erythrocytes drawn from allyl alcohol-intoxicated mice. The release is accompanied by peroxidation of membrane lipids when the cells are depleted of glutathione. Lipid peroxidation is always followed by the lysis of the cells. The release of iron is also accompanied by methemoglobin formation, but the extent of the release does not correlate with the level of methemoglobin production. The addition of DFO to the incubation mixture or the preincubation of the erythrocytes with DFO in millimolar concentrations completely prevents both lipid peroxidation and hemolysis while not significantly changing the level of iron release. Morphological studies carried out with scanning electron microscopy showed a number of alterations in the shape of the incubated erythrocytes, including echinocyte transformation and the appearance of codocyte, stomatocyte, and cnizocyte like forms. These alterations were more prominent with increasing lipid peroxidation and hemolysis, even if occurring in their absence. On the contrary, the appearance of pits and holes was strictly associated with lipid peroxidation and lysis.

Iron released from an erythrocyte lysate by oxidative stress is diffusible and in redox active form.

The incubation of a ghost-free erythrocyte lysate with the oxidizing agent phenylhydrazine resulted in both methemoglobin formation and release of iron in a desferrioxamine (DFO)-chelatable form. The released iron was diffusible, as shown by a dialysis carried out simultaneously with the incubation. When the dialysate was added to erythrocyte ghosts or to microsomes from liver or brain, lipid peroxidation developed in the membranes, indicating that the diffusible iron was in a redox active form. The addition of ATP to the lysate markedly increased both iron diffusion and lipid peroxidation in the membranes subsequently added to the dialysate. The possible implication of these data in some well known pathologies is discussed.

Iron release and membrane damage in erythrocytes exposed to oxidizing agents, phenylhydrazine, divicine and isouramil.

Mouse erythrocytes were incubated with oxidizing agents, phenylhydrazine, divicine and isouramil. With all the oxidants a rapid release of iron in a desferrioxamine (DFO)-chelatable form was seen and it was accompanied by methaemoglobin formation. If the erythrocytes were depleted of GSH by a short preincubation with diethyl maleate, the release of iron was accompanied by lipid peroxidation and, subsequently, haemolysis. GSH depletion by itself did not induce iron release, methaemoglobin formation, lipid peroxidation or haemolysis. Rather, the fate of the cell in which iron is released depended on the intracellular availability of GSH. In addition, iron release was higher in depleted cells than in native ones, suggesting a role for GSH in preventing iron release when oxidative stress is imposed by the oxidants. Iron release preceded lipid peroxidation. The latter was prevented when the erythrocytes were preloaded with DFO in such a way (preincubation with 10 mM-DFO) that the intracellular concentration was equivalent to that of the released iron, but not when the intracellular DFO was lower (preincubation with 0.1 mM-DFO). Extracellular DFO did not affect lipid peroxidation and haemolysis, suggesting again that the observed events occur intracellularly (intracellular chelation of released iron). The relevance of iron release from iron complexes in the mechanisms of cellular damage induced by oxidative stress is discussed.

Iron release and erythrocyte damage in allyl alcohol intoxication in mice.

Allyl alcohol administration in a toxic dose (1.5 mmol/kg) to starved mice causes the development of hemolysis in nearly 50% of the animals. Malonic dialdehyde (MDA) appears in plasma of the animals showing hemolysis. The treatment of mice with desferrioxamine after allyl alcohol intoxication completely prevents lipid peroxidation and hemolysis, suggesting the involvement of iron in the allyl alcohol-induced erythrocyte damage. Erythrocytes obtained from intoxicated mice before the development of hemolysis show, upon incubation, release of iron, lipid peroxidation and lysis. Studies carried out with reconstituted systems of erythrocyte lysates, containing ghosts and different fractions of erythrocyte cytosol and incubated in the presence of acrolein (the major metabolite of allyl alcohol), strongly suggest that iron is released from hemoglobin. This iron appears to promote lipid peroxidation which is accompanied by erythrocyte lysis. Thus, the allyl alcohol-induced hemolysis appears to be a model for iron delocalization from iron stores.

Increased prooxidant action of hepatic cytosolic low-molecular-weight iron in experimental iron overload.

In the iron-loaded liver there may be an increase in the putative intracellular transit pool of iron, components of which could be catalytically active in stimulating lipid peroxidation. To study the levels of low-molecular-weight, catalytically active iron in the liver, cytosolic ultrafiltrates were tested in an assay containing rat liver microsomes and NADPH. Malondialdehyde production was used as an index of lipid peroxidation. This assay system was sensitive enough to detect 0.25 mumol/L ferrous iron; progressive but non-linear increases in malondialdehyde were produced as the iron concentration was increased to 5 mumol/L. Ultrafiltrates from hepatic cytosol of iron-loaded rats had greater prooxidant action than did those from controls. When added to the assay, deferoxamine, an iron chelator, completely suppressed the prooxidant action of hepatic ultrafiltrates, showing that this activity is iron-dependent. Deferoxamine administered intraperitoneally to control animals at a dose of 1 gm/kg completely inhibited the prooxidant effect of hepatic ultrafiltrates prepared from rats killed after 1, 2 and 3 hr. Partial inhibition was observed at 4 hr; by 6 hr the inhibitory effect of deferoxamine was completely lost. Administration of deferoxamine (1 gm/kg intraperitoneally, 1 hr before killing) completely inhibited the prooxidant action of hepatic ultrafiltrates in moderately iron-loaded rats and controls but had no protective effect in heavily iron-loaded rats. These results support the concept that iron overload results in an increase in a hepatic cytosolic pool of low-molecular-weight iron that is catalytically active in stimulating lipid peroxidation. This pool can be chelated transiently in vivo by deferoxamine in moderate, but not heavy, iron overload.

Allyl alcohol-induced hemolysis and its relation to iron release and lipid peroxidation.

Allyl alcohol administration to starved mice produced, along with liver necrosis, a high incidence (about 50%) of hemolysis. A marked decrease in erythrocyte glutathione (GSH) was seen in all the intoxicated animals. Such a decrease was significantly higher in the animals showing hemolysis. In these animals a substantial amount of malonic dialdehyde (MDA) was detected in plasma and a marked decrease in arachidonic and docosahexaenoic acids was found in erythrocyte phospholipids. These data suggest that the allyl alcohol-induced hemolysis is mediated by lipid peroxidation. In vitro studies have shown that the addition of acrolein to mouse erythrocytes produces a dramatic GSH depletion, which is followed by the appearance of lipid peroxidation and, after an additional 30 min of incubation, by the development of hemolysis. Prevention of lipid peroxidation by an antioxidant (Trolox C) or an iron chelator (desferrioxamine, DFO), prevented hemolysis even if the erythrocyte GSH level was dramatically decreased. In vitro, allyl alcohol and acrylic acid were ineffective in inducing GSH depletion, lipid peroxidation and hemolysis. Studies of possible induction of lipid peroxidation in erythrocytes showed that a progressive increase in "free" (desferal chelatable) iron occurs in the erythrocytes during the incubation with acrolein. It seems, therefore, that a release of iron from iron-containing complexes occurs in acrolein-treated erythrocytes and that such "free" iron promotes lipid peroxidation.

Lipid peroxidation, protein thiols and calcium homeostasis in bromobenzene-induced liver damage.

The mechanisms of bromobenzene hepatotoxicity in vivo were studied in mice. The relationships among glutathione (GSH) depletion, lipid peroxidation, loss of protein thiols, disturbed calcium homeostasis and liver necrosis were investigated. Liver necrosis (as estimated by the serum glutamate-pyruvate transaminase (SGPT) level) appeared between 9 and 12 hr and increased at 18 hr. Lipid peroxidation which was already detectable at 6 hr in some animals, increased thereafter showing a good correlation with the severity of liver necrosis. Despite a quite fast depletion of hepatic GSH, a significant decrease in protein thiols could be observed at 12-18 hr only. Loss of protein thiols in both whole liver and subcellular fractions (microsomes and mitochondria) was correlated with lipid peroxidation. Also a good inverse correlation was seen between lipid peroxidation and the calcium sequestration activity of liver microsomes and mitochondria. The treatment of mice with desferrioxamine (DFO) after bromobenzene-intoxication completely prevented lipid peroxidation, loss of protein thiols and liver necrosis in the animals sacrificed 15 hr after poisoning. When, however, the animals were examined at 24 hr, although the general correlation between lipid peroxidation and liver necrosis was held, in some animals (about 30% of the survivors) elevation of SGPT was observed in the virtual absence of lipid peroxidation. It seems likely therefore that the liver damage seen during the first phase of bromobenzene-intoxication is strictly related to lipid peroxidation. It is, however, possible that in some animals in which for some reason lipid peroxidation does not develop, another mechanism of liver necrosis unrelated to lipid peroxidation occurs at later times.