Iron-sulfur proteins are the major source of protein-bound dinitrosyl iron complexes formed in Escherichia coli cells under nitric oxide stress

Protein-bound dinitrosyl iron complexes (DNICs) have been observed in prokaryotic and eukaryotic cells under nitric oxide (NO) stress. The identity of proteins that bind DNICs, however, still remains elusive. Here we demonstrate that iron-sulfur proteins are the major source of protein-bound DNICs formed in Escherichia coli cells under NO stress. Expression of recombinant iron-sulfur proteins, but not proteins without iron-sulfur clusters, almost doubles the amount of protein-bound DNICs formed in E. coli cells after NO exposure. Purification of recombinant proteins from the NO-exposed E. coli cells further confirms that iron-sulfur proteins, but not proteins without iron-sulfur clusters, are modified, forming protein-bound DNICs. Deletion of the iron-sulfur cluster assembly proteins IscA and SufA to block the [4Fe-4S] cluster biogenesis in E. coli cells largely eliminates the NO-mediated formation of protein-bound DNICs, suggesting that iron-sulfur clusters are mainly responsible for the NO-mediated formation of protein-bound DNICs in cells. Furthermore, depletion of the "chelatable iron pool" in wild-type E. coli cells effectively removes iron-sulfur clusters from proteins and concomitantly diminishes the NO-mediated formation of protein-bound DNICs, indicating that iron-sulfur clusters in proteins constitute at least part of the chelatable iron pool in cells.     

Reactivity of nitric oxide with the [4Fe-4S] cluster of dihydroxyacid dehydratase from Escherichia coli

Although the NO (nitric oxide)-mediated modification of iron-sulfur proteins has been well-documented in bacteria and mammalian cells, specific reactivity of NO with iron-sulfur proteins still remains elusive. In the present study, we report the first kinetic characterization of the reaction between NO and iron-sulfur clusters in protein using the Escherichia coli IlvD (dihydroxyacid dehydratase) [4Fe-4S] cluster as an example. Combining a sensitive NO electrode with EPR (electron paramagnetic resonance) spectroscopy and an enzyme activity assay, we demonstrate that NO is rapidly consumed by the IlvD [4Fe-4S] cluster with the concomitant formation of the IlvD-bound DNIC (dinitrosyl-iron complex) and inactivation of the enzyme activity under anaerobic conditions. The rate constant for the initial reaction between NO and the IlvD [4Fe-4S] cluster is estimated to be (7.0+/-2.0)x10(6) M(-2) x s(-1) at 25 degrees C, which is approx. 2-3 times faster than that of the NO autoxidation by O2 in aqueous solution. Addition of GSH failed to prevent the NO-mediated modification of the IlvD [4Fe-4S] cluster regardless of the presence of O2 in the medium, further suggesting that NO is more reactive with the IlvD [4Fe-4S] cluster than with GSH or O2. Purified aconitase B [4Fe-4S] cluster from E. coli has an almost identical NO reactivity as the IlvD [4Fe-4S] cluster. However, the reaction between NO and the endonuclease III [4Fe-4S] cluster is relatively slow, apparently because the [4Fe-4S] cluster in endonuclease III is less accessible to solvent than those in IlvD and aconitase B. When E. coli cells containing recombinant IlvD, aconitase B or endonuclease III are exposed to NO using the Silastic tubing NO delivery system under aerobic and anaerobic conditions, the [4Fe-4S] clusters in IlvD and aconitase B, but not in endonuclease III, are efficiently modified forming the protein-bound DNICs, confirming that NO has a higher reactivity with the [4Fe-4S] clusters in IlvD and aconitase B than with O2 or GSH. The results suggest that the iron-sulfur clusters in proteins such as IlvD and aconitase B may constitute the primary targets of the NO cytotoxicity under both aerobic and anaerobic conditions.     

Protein-bound dinitrosyl-iron complexes appearing in blood of rabbit added with a low-molecular dinitrosyl-iron complex: EPR studies.

The formation of protein-bound dinitrosyl-iron complexes (DNIC) in blood plasma and packed red cell fraction has been demonstrated by the EPR method in the experiments on rabbits which were i/v injected with the low-molecular DNIC with thiosulphate. This formation was ensured by transfer of Fe(+)(NO(+))(2) moieties from low-molecular DNIC onto serum albumin or hemoglobin molecules. Protein-bound DNICs appeared immediately after low-molecular DNIC injection followed with gradually decreasing their amounts. The complexes could be detected by EPR technique during more than two days. The addition of water-soluble NO scavenger, the iron complex with N-methyl-d-glucamine dithiocarbamate (MGD) resulted in decomposition of a part of protein-bound DNICs and in effective excretion of secondary products (mainly mononitrosyl-iron complexes with MGD) from the blood flow.     

Distribution and pharmacokinetics of dinitrosyl iron complexes in rat organs

Dinitrosyl iron complexes (DNICs) have been traced in rat blood and organs after intravenous infusion of Oxacom. It is shown that the active principle (DNIC with glutathione) is rapidly distributed through the organism and deposited in blood and organs as protein-bound DNICs. The specific levels of DNIC in the main body organs are comparable, whereas its apparent lifetimes relate as blood < heart = lung < liver < kidney. Spin trapping assays indicate that protein-bound DNICs are a major but not the only form of NO deposition; the next largest depot is most probably formed by S-nitrosothiols. The gradual release of NO from such pools ensures the smooth and prolonged hypotensive effect of Oxacom.     

Long-lasting hypotensive action of stable preparations of dinitrosyl-iron complexes with thiol-containing ligands in conscious normotensive and hypertensive rats

Previously we established the hypotensive action of nitric oxide donors, dinitrosyl-iron complexes (DNIC) with thiol-containing ligands, stored in frozen solution at 77K. In the present study, we tested recently designed water soluble dry powder preparations of DNICs keeping their characteristics in dry air for a long time. The complexes dissolved in PBS were injected intravenously into normotensive Wistar and spontaneously hypertensive SHR rats. The average arterial pressure (AAP) was recorded through preliminary implanted catheter in a carotid artery. The initial hypotensive action of DNIC with cysteine (DNIC-cys) was comparable to action of nitroprusside (SNP) but, in contrast to the latter, lasted for 20-120min depending on a doze. The blood DNIC content as detected by electronic paramagnetic resonance steadily decreased at this time. The hypotensive action of S-nitrosocysteine was similar to SNP while binding of iron in DNIC by batophenantroline-disulphonate prevented its hypotensive effect. These data suggest that long-lasting hypotensive action of DNICs may be caused by stable protein-bound DNICs forming in the process of transfer of Fe(+)(NO(+))(2) moieties from low-molecular DNICs to thiol protein ligands. The relative initial dose-dependent effect of DNIC-cys was similar in Wistar and SHR but secondary AAP reduction was more profound in SHR. A substitution of cysteine in DNIC by thiosulphate resulted in markedly less initial AAP reduction while long-lasting effect was similar and substitution by glutathione smoothed initial AAP decline and stabilized AAP level in the second phase. Prolonged AAP reduction induced by DNIC-cys was considerably shortened in narcotized rats. Thus, dry preparations of DNICs preserve prolonged hypotensive activity.     

Study of dinitrosyl-iron complexes pharmacokinetics and accumulation in depot in rat organs

Protein-bound dinitrosyl-iron complexes (DNIC) in rat whole blood and organs were studied after intravenous injection of this substance with glutathione ligand (DNIC-GH). The effect of DNIC-GH injection on NO level (including NO physiological forms) in hydrophobic areas of rat tissues was also studied in normal physiological blood circulation condition. It has been shown, that after DNIC-GH injection the concentration of protein-bound DNICs in rat whole blood and organs rapidly reached maximum values, and then gradually decreased, that pointed to decomposition of DNIC molecules, coupled with NO release. At the beginning of the experiment the rates of DNIC decay in rat heart and lung were substantially higher, as compared with those in liver and kidney. By spin trappping it has been demonstrated that DNIC-GH, as a source of NO physiological forms (including S-nitrosothiols), in normal physiological blood circulation influence heart more selectively, as compared with the other organs.     

Dinitrosyl iron complexes--structure and biological functions

Formation of dinitrosyl iron complexes (DNICs), which can be described by general formula Fe(NO)2(L)2, where L is carbonyl-, nitrosyl- or imino- complexing ligand, was observed in many kinds of living organisms, in a wide spectrum of physiological conditions associated with inflammation, ischemia/reperfusion and cancer. Accumulation of DNICs coincides with intensified production of nitric oxide in macrophages, neurons, endothelial cells, Langerhans' cells and hepatocytes. Low-molecular thiol-containing DNICs (DNIC-(RS)2) show vasodilatory action and they are proposed to play a role of nitric oxide transducers and stabilizers. DNICs have been shown to modulate redox potential of the cell via inhibition of glutathione-dependent enzymes, such as glutathione reductase, S-transferase and peroxidase. Although there is a convincing experimental evidence for their NO and NO+ donating function, the nature of DNICs formed in biological systems, their stability and biological role is still a matter of discussion.     

Nitric oxide storage and transport in cells are mediated by glutathione S-transferase P1-1 and multidrug resistance protein 1 via dinitrosyl iron complexes

Nitrogen monoxide (NO) plays a role in the cytotoxic mechanisms of activated macrophages against tumor cells by inducing iron release. We showed that NO-mediated iron efflux from cells required glutathione (GSH) (Watts, R. N., and Richardson, D. R. (2001) J. Biol. Chem. 276, 4724-4732) and that the GSH-conjugate transporter, multidrug resistance-associated protein 1 (MRP1), mediates this release potentially as a dinitrosyl-dithiol iron complex (DNIC; Watts, R. N., Hawkins, C., Ponka, P., and Richardson, D. R. (2006) Proc. Natl. Acad. Sci. U.S.A. 103, 7670-7675). Recently, glutathione S-transferase P1-1 (GST P1-1) was shown to bind DNICs as dinitrosyl-diglutathionyl iron complexes. Considering this and that GSTs and MRP1 form an integrated detoxification unit with chemotherapeutics, we assessed whether these proteins coordinately regulate storage and transport of DNICs as long lived NO intermediates. Cells transfected with GSTP1 (but not GSTA1 or GSTM1) significantly decreased NO-mediated 59Fe release from cells. This NO-mediated 59Fe efflux and the effect of GST P1-1 on preventing this were observed with NO-generating agents and also in cells transfected with inducible nitric oxide synthase. Notably, 59Fe accumulated in cells within GST P1-1-containing fractions, indicating an alteration in intracellular 59Fe distribution. Furthermore, electron paramagnetic resonance studies showed that MCF7-VP cells transfected with GSTP1 contain significantly greater levels of a unique DNIC signal. These investigations indicate that GST P1-1 acts to sequester NO as DNICs, reducing their transport out of the cell by MRP1. Cell proliferation studies demonstrated the importance of the combined effect of GST P1-1 and MRP1 in protecting cells from the cytotoxic effects of NO. Thus, the DNIC storage function of GST P1-1 and ability of MRP1 to efflux DNICs are vital in protection against NO cytotoxicity.     

Formation of dinitrosyl iron complexes in cardiac mitochondria

It has been established that, in the presence of S-nitrosothiols, cysteine, and mitochondria, dinitrosyl iron complexes (DNIC) coupled to low-molecular-weight ligands and proteins are formed. The concentration of DNIC depended on oxygen partial pressure. It was shown that, under the conditions of hypoxia, the kinetics of the formation of low-molecular DNIC was biphasic. After the replacement of anaerobic conditions of incubation with aerobic ones, the level of DNIC came down; in this case, protein dinitrosyl complexes became more stable. We proposed that iron-and sulfur-containing proteins and low-molecular-weight iron complexes are the sources of iron for DNIC formation in mitochondrial suspensions. It was shown that a combination of DNIC and S-nitrosothiols inhibited effectively the respiration of cardiomyocytes.     

In vivo evidence for the iron-binding activity of an iron-sulfur cluster assembly protein IscA in Escherichia coli

IscA is a key member of the iron-sulfur cluster assembly machinery in prokaryotic and eukaryotic organisms; however, the physiological function of IscA still remains elusive. In the present paper we report the in vivo evidence demonstrating the iron-binding activity of IscA in Escherichia coli cells. Supplement of exogenous iron (1 μM) in M9 minimal medium is sufficient to maximize the iron binding in IscA expressed in E. coli cells under aerobic growth conditions. In contrast, IscU, an iron-sulfur cluster assembly scaffold protein, or CyaY, a bacterial frataxin homologue, fails to bind any iron in E. coli cells under the same experimental conditions. Interestingly, the strong iron-binding activity of IscA is greatly diminished in E. coli cells under anaerobic growth conditions. Additional studies reveal that oxygen in medium promotes the iron binding in IscA, and that the iron binding in IscA in turn prevents formation of biologically inaccessible ferric hydroxide under aerobic conditions. Consistent with the differential iron-binding activity of IscA under aerobic and anaerobic conditions, we find that IscA and its paralogue SufA are essential for the iron-sulfur cluster assembly in E. coli cells under aerobic growth conditions, but not under anaerobic growth conditions. The results provide in vivo evidence that IscA may act as an iron chaperone for the biogenesis of iron-sulfur clusters in E. coli cells under aerobic conditions.     

Regulation of the citrate synthase (gltA) gene of Escherichia coli in response to anaerobiosis and carbon supply: role of the arcA gene product.

As an enzyme of the tricarboxylic acid cycle pathway, citrate synthase participates in the generation of a variety of cellular biosynthetic intermediates and in that of reduced purine nucleotides that are used in energy generation via electron transport-linked phosphorylation reactions. It catalyzes the condensation of oxaloacetate and acetyl coenzyme A to produce citrate plus coenzyme A. In Escherichia coli this enzyme is encoded by the gltA gene. To investigate how gltA expression is regulated, a gltA-lacZ operon fusion was constructed and analyzed following aerobic and anaerobic cell growth on various types of culture media. Under aerobic culture conditions, expression was elevated to a level twofold higher than that reached under anaerobic culture conditions. ArcA functions as a repressor of gltA expression under each set of conditions: in a delta arcA strain, gltA-lacZ expression was elevated to levels two- and eightfold higher than those seen in a wild-type strain under aerobic and anaerobic conditions, respectively. This control is independent of the fnr gene product, an alternative anaerobic gene regulator in E. coli. When the richness or type of carbon compound used for cell growth was varied, gltA-lacZ expression varied by 10- to 14-fold during aerobic and anaerobic growth. This regulation was independent of both the crp and fruR gene products, suggesting that another regulatory element in E. coli is responsible for the observed control. Finally, gltA-lacZ expression was shown to be inversely proportional to the cell growth rate. These findings indicate that the regulation of gltA gene expression is complex in meeting the differential needs of the cell for biosynthesis and energy generation under various cell culture conditions.     

The source of the iron binding with nitric oxide in activated macrophages]

No decrease in iron-sulphur centers was found in cultured macrophage cells (J774) after the treatment with nitric oxide (10(-7) M NO/10(7) cells) during 5 min. The center content was controlled by the electron spin resonance (ESR) method. The macrophages pretreated with dithionite + methyl viologen showed the formation of dinitrosyl iron complexes (DNIC) with a characteristic ESR signal at g approximately 2.03. The data suggest that loosely bound nonheme iron (free iron) mostly contributes to the formation of these complexes. Iron from iron-containing proteins does not release from these centers under the direct action of nitric oxide. The iron-sulphur centers can be destroyed by the products of nitric oxide oxidation (NO2, N2O3, etc.) as oxidizing and acid agents.     

How does oxygen inhibit central metabolism in the obligate anaerobe Bacteroides thetaiotaomicron

The molecular basis of obligate anaerobiosis is not well established. Bacteroides thetaiotaomicron is an opportunistic pathogen that cannot grow in fully aerobic habitats. Because microbial niches reflect features of energy-producing strategies, we suspected that aeration would interfere with its central metabolism. In anaerobic medium, this bacterium fermented carbohydrates to a mixture of succinate, propionate and acetate. When cultures were exposed to air, the formation of succinate and propionate ceased abruptly. In vitro analysis demonstrated that the fumarase of the succinate-propionate pathway contains an iron-sulphur cluster that is sensitive to superoxide. In vivo, fumarase activity fell to < 5% when cells were aerated; virtually all activity was recovered after extracts were chemically treated to rebuild iron-sulphur clusters. Aeration minimally affected the remainder of this pathway. However, aeration reduced pyruvate:ferredoxin oxidoreductase (PFOR), the first enzyme in the acetate fermentation branch, to 3% of its anaerobic activity. This cluster-containing enzyme was damaged in vitro by molecular oxygen but not by superoxide. Thus, aerobic growth is precluded by the vulnerability of these iron-sulphur cluster enzymes to oxidation. Importantly, both enzymes were maintained in a stable, inactive form for long periods in aerobic cells; they were then rapidly repaired when the bacterium was returned to anaerobic medium. This result explains how this pathogen can easily recover from occasional exposure to oxygen.     

Effect of growth conditions on catabolite repression and cyclic AMP synthesis in Escherichia coli 3000A1

Catabolite repression of beta-galactosidase synthesis in E. coli 3000A1 (adenine-) was studied under a variety of growth conditions. The differential rate of induced beta-galactosidase synthesis was maximal at the growth rate of 0.75 division per h, irrespective of whether growth conditions were aerobic or anaerobic. The addition of cyclic AMP (cAMP) to the medium partly restored the repressed synthesis of beta-galactosidase under some growth conditions, but showed little or no effect on the enzyme synthesis under other conditions. Although growth rate and profile of beta-galactosidase synthesis in glucose-grown cells were similar to those in arabinose-grown cells, the acceleration of beta-galactosidase synthesis upon the addition of cAMP was found only in glucose-grown cells. The cells aerobically grown in the presence of glycerol, xylose, or arabinose showed a high synthetic rate of cAMP and were insensitive to exogenously supplied cAMP as regards beta-galactosidase synthesis. Although the cells grown with glucose showed similar rates of cAMP synthesis under aerobic and anaerobic conditions, the differential rate of beta-galactosidase synthesis was much higher in the anaerobic state than in the aerobic state. These findings support the idea that catabolite repression found in the strain is caused through two mechanisms, i.e., cAMP-mediated and cAMP-independent ones.