In 1964–1965, R.M. Nalbandyan and A.F. Vanin [1, 2] discovered paramagnetic centers with a characteristic EPR signal with g factor values \({{g}_{ \bot }}\) = 2.04 and \({{g}_{\parallel }}\) = 2.014 in yeast cells. This signal was denoted as 2.03 according to its median value. In 1967, they were found to belong to the mononuclear form of iron dinitrosyl complexes (M-DNIC) with thiol-containing ligands [3]. In the same year, these complexes were detected by the 2.03 signal in animal tissues [4]. Later, our team, headed by A.F. Vanin, not only determined their composition and origin but also demonstrated their diverse biological activities, stemming from their ability to donate nitric oxide, one of the universal agents that regulate metabolic processes in all living species [5, 6].

It has been found that diamagnetic (EPR-silent) binuclear forms of DNICs with thiol-containing ligands (B-DNICs) may arise in various living systems along with the corresponding mononuclear forms [5–10]. We studied the electron and spatial structures of both DNIC forms and their ability to act as donors of not only neutral NO molecules but also nitrosonium cations (NO+), was no less active biologically. We proposed M- and B‑DNICs formulas that characterize one of their major resonance structures as [(RS–)2Fe+(NO+)2] and [(RS–)2(Fe+)2(NO+)4], respectively [5, 6].

In addition, there is reason to believe that DNICs with thiol-containing ligands that form in various living species that produce nitric oxide are a working form of this agent; that is, NO can exert its physiological action on human and animal bodies only being incorporated in a DNIC as either NO or NO+. DNICs support the deposition and stabilization of these agents, as well as their delivery to targets: heme- and thiol-containing proteins, respectively. The subsequent binding of NO and NO+ to the heme and thiol groups of proteins generates nitrosyl complexes of the heme iron and S-nitrosothiols, respectively, thereby dramatically altering the conformation of the protein involved to activate or deactivate it [5, 6].

Recently, we found evidence that DNICs with thiol-containing ligands present in animal tissues exist mainly in the binuclear form [7, 8]. These data were obtained in experiments with treatment of isolated animal organs with dithiocarbamate derivatives: diethyl dithiocarbamage (DETC) or N-methyl-D-glucamine dithiocarbamate (MGD). As shown in Fig. 1, these derivatives can take a mononitrosyl iron fragment from the dinitrosyl iron group of B-DNIC to produce EPR-active mononitrosyl iron complexes (MNICs) with DETC or MGD and S-nitrosothiols RS–NO+ [7, 8]. Being water-insoluble, MNIC−DETC resides mainly in hydrophobic membrane compartments, whereas water-soluble MNIC−MGD resides in water-bearing compartments of cells and tissues.

Fig. 1.
figure 1

Conversion of M- and B-DNICs with thiol-containing ligands to MNIC with diethyldithiocarbamate.

Full size image

Our previous study was dedicated to the interaction of DETC and MGD with isolated organs of mice that had been intraperitoneally administered exogenous NO donors: sodium nitrite, organic nitrite isosorbide dinitrate (ISDN), or S-nitrosoglutathione [8]. As well, some organs were treated in vitro with gaseous NO and then incubated in DETC or MGD solutions. In experiments with endogenous NO, its formation was mediated by inflammation induced by administration of bacterial lipopolysaccharides. As a result, the mouse tissues intensely produced the potent NO generator inducible NO synthase (iNOS). The high NO level in the liver was assessed from the amount of MNIC−DETC formed in vitro as a result of NO trapping by iron complexes with DETC. This level was compared with the level of MNIC−DETC formed in vitro in mice liver treated with lipopolysaccharides after tissue incubation in DETC solution.

In [7], isolated liver was treated with DETC solutions by perfusion. It was shown that intraperitoneal injection of B-DNIC with glutathione caused the formation of large MNIC−DETC levels in the liver, four to fivefold greater than the level of M-DNIC detected in liver in vivo after administration of B-DNIC with glutathione (B-DNIC−GSH).

This raises the question of whether the administration of all mentioned NO donors to mice caused in all cases the emergence of diamagnetic B-DNICs with thiol-containing ligands, which then were converted by DETC (or MGD) to paramagnetic EPR-active MNIC–DETC (MGD) complexes. This experimental work is an attempt to answer this question.

MATERIALS AND METHODS

Materials. Ferrosulfate (FeSO4 · 7H2O) was purchased from Fluka (Switzerland); reduced glutathione (GSH), sodium dithionite, N-acetylcysteine(NAC), and sodium citrate, from Sigma (United States); isosorbide dinitrate (ISDN, Isoket®) from UCB Pharma (Ireland); and DETC from Acros (United States). Gaseous NO was obtained by the reaction of ferrosulfate with sodium nitrite in 0.1 M HCl followed by purification from NO2 by cryogenic sublimation in vacuo [9].

Syntheses of B-DNIC with glutathione, B-DNIC with N-acetyl-L-cysteine, and S-nitrosoglutathione. B‑DNIC−GSH was obtained as in [10]. Glutathione (120 mg, 40 mM), ferrosulfate (56 mg, 20 mM, and sodium nitrite (14 mg) were sequentially diluted in 10 mL of 15 mM HEPES buffer pH 7.4. The addition of glutathione decreased the pH of the solution considerably, to 3.5, and ferrosulfate was dissolved without forming water-insoluble iron hydroxide complexes; thus, GS−NO formed after the addition of nitrite to stain the solution pink. At 1 to 1.5 hours after the completion of GS-NO synthesis (noted from GS–NO absorbance at 334 nm), pH of the solution was adjusted to neutral by dropwise addition of 0.1 M NaOH. The color of the solution changed from pink to dark orange, characteristic of the B-DNIC complex with glutathione (B-DNIC−GSH). The ionized iron not incorporated in the complex gradually precipitated in the form of iron hydroxide complexes.

The resulting suspension was left to stay in the air at room temperature for 8–12 h until complete precipitation of iron hydroxide complexes, which were then removed from the solution by paper filtration. The concentration of B-DNIC−GSH was determined from the intensities of the characteristic bands at 310 and 360 nm, whose molar attenuation coefficients are 4600 and 3700 M–1 cm–1, respectively (per one iron atom in B-DNIC) [9]. The concentration of the synthesized B-DNIC−GSH averaged 9.5 mM.

The complex of B-DNIC with N-acetyl-L-cysteine (B-DNIC−NAC) was synthesized in a similar way with some amendments. The synthesis of S-nitroso-NAC took no more than 20 min, and of B-DNIC−NAC took no more than 3–4 h. In some experiments we used B-DNIC−GSH/NAC synthesized by treatment of Fe2+ with gaseous NO followed by reaction of the solutions with solutions of GSH or NAC as described in [9].

To prepare S-nitrosoglutathione (GS−NO), we prepared first 5.5 mM glutathione solution in 15 mM HEPES. After GSH addition, the pH of the buffer decreased to 3.0–3.5. We added 5 mmol of sodium nitrite to the solution in the air. The solution color became pink, as is characteristic of GS−NO. The product concentration was assessed from its absorbance at 334 nm, where the molar attenuation coefficient is 0.9 M–1 cm–1 [11].

Animal studies. Experiments were conducted with outbred laboratory male mice weighing 18–20 g. All substances except for the iron citrate complex were injected intraperitoneally in 0.2 mL of normal saline at the following doses: NaNO2, 50 mg/kg (0.7 mM/kg); ISDN, 1 mM/kg; DETC, 500 mg/kg (3 mM/kg); GS−NO, 0.2 mM/kg; B-DNIC−GSH/NAC, 0.12–0.15 mM/kg (per one iron atom in B-DNIC). Iron citrate was injected intramuscularly into a paw at the following doses: FeSO4 · 7H2O, 37.5 mg (0.135 mM/kg); citrate, 187.5 mg (0.675 mM/kg).

The animals treated with NaNO2, ISDN, GS−NO, or B-DNIC−GSH with synchronous administration of the iron citrate complex were decapitated 30 min after the procedure. The livers were extracted, placed to EPR tubes, and frozen in liquid nitrogen. After recording EPR spectra, the samples were thawed and kept in 200 mM DETC solution at room temperature for 30 min. The samples were then frozen again and EPR spectra were recorded.

The EPR spectra of liver samples were recorded with a modified Radio-Pan spectrometer (Poland) of the X range at 77 K in cylindrical EPR tubes 4 mm in diameter placed to an insert of a quartz vacuum flask. Paramagnetic particle concentrations were determined by double integration of EPR signals with reference to a frozen M-DNIC−GSH solution of known concentration.

Statistical evaluation. Experimental results are presented as the mean values ± the standard error of the mean from three or four measurements.

RESULTS

MNIC−DETC formation in liver tissue of animals treated with B-DNIC−NAC or B-DNIC−GSH. Intraperitoneal injection of 0.12–0.15 mM/kg B-DNIC–NAC to mice resulted in the appearance of M-DNICs in their livers. It was apparent from the amplitude of the characteristic 2.03 signal that their concentrations varied little if at all within 1 h after B-DNIC administration, being 90 ± 20 μmol/kg wet liver tissue. Additional intraperitoneal injection of 500 mg/kg (3 mM/kg) of DETC 30 min after the B-DNIC−NAC injection resulted in complete disappearance of the 2.03 signal for 30 min. Instead, an intense EPR signal of MNIC−DETC with \({{g}_{ \bot }}\) = 2.035, \({{g}_{\parallel }}\) = 2.02, and triplet hyperfine coupling at \({{g}_{ \bot }}\) (Fig. 2e). Judging from this signal, the MNIC−DETC concentration significantly exceeded the M-DNIC concentration, being 400 ± 30 μmol/kg wet liver tissue.

Fig. 2.
figure 2

EPR spectra recorded in mouse liver samples: (a, b) 30 min after intraperitoneal injection of B-DNIC−NAC or DNIC−GSH; (c, d) 30 min after incubation of liver samples from identical mice in 200 mM DETC; (e, f) in liver samples from mice receiving the same treatment followed by intraperitoneal DETC injection 30 min before sacrificing. The spectra were recorded at 77 K. On the right: radiospectrometer amplification, a.u.

Full size image

A different situation arose in experiments where liver samples were taken 30 min after B-DNIC–NAC injection and kept in the air in 15 mM HEPES buffer pH 7.4 with 200 mM DETC. This incubation also resulted in M-DNIC disappearance from liver samples and appearance of MNIC–DETC (Fig. 2c). In this case, however, the concentration of the latter did not exceed that of the initial M-DNICs. Addition of the pulverized strong reducer sodium dithionite did not affect the MNIC–DETC concentration (data not shown).

Similar experiments with animals where B-DNIC−NAC was replaced with B-DNIC−GSH yielded similar results (Fig. 2, right panel). The only difference was that the levels of M-DNIC and MNIC−DETC that formed after intraperitoneal injection of B-DNIC−GSH and DETC at the same doses as with B-DNIC−NAC were much lower: 15 ± 5 and 80 ± 15 μmol/kg wet tissue, respectively (Figs. 2b and 2f). These lower concentrations may be explained by lower permeation of B-DNIC−GSH from the peritoneum to blood vessels, probably because of the much lower glutathione transmission through cell membranes as compared to the extremely efficient NAC transmission. After incubation of isolated liver samples in 200 mM DETC, the concentration of the formed MNIC−DETC was 20 ± 5 μmol/kg wet tissue (Fig. 2d).

Use of either appropriate S-nitrosothiols (GS−NO or S-nitroso-NAC) or gaseous NO in the synthesis of B-DNICs made no differences in experimental results.

MNIC−DETC formation in mouse liver tissue treated in vitro with gaseous NO and diethyldithiocarbamate. As in our earlier study [8], successive treatment of isolated mouse liver tissue with NO and DETC by 30-min incubation in 200 mM DETC at neutral pH produced MNIC−DETC, whose concentration was 250 ± 50 μmol/kg wet tissue on the average, whereas the average M-DNIC concentration was 40 ± 10 μmol/kg wet tissue (Figs. 3a and 3c). It is worth noting that in this case treatment of liver tissues containing MNIC−DETC with sodium dithionite did not affect its level either (Fig. 3e). As in [8], the high level of MNIC−DETC mentioned above was observed in liver samples from mice that received intraperitoneally iron citrate complex prior to sacrificing. Liver samples from animals that had not received this treatment showed three to fivefold lesser MNIC−DETC concentrations.

Fig. 3.
figure 3

EPR spectra of liver samples from mice: left: (a) treated in vitro with gaseous NO at 150 mmHg; (c) with subsequent 30-min incubation of the samples in 200 mM DETC; (e) treated then with dithionite; right: (b) 30 min after intraperitoneal injection of GS−NO; (d) with subsequent 30-min incubation in 200 mM DETC; (f) treated then with dithionite. The spectra were recorded at 77 K.

Full size image

In vitro treatment of liver samples with gaseous NO caused the formation of not only DNIC but also nitrosyl complexes of heme-containing proteins (mainly hemoglobin), characterized by an EPR signal with gi values 2.07 and 1.98 (Fig. 3a).

MNIC−DETC formation in liver tissue of animals treated in vivo with S-nitrosoglutathione. The 2.03 signal was recorded within 30 min in the livers of mice that received GS-NO and the Fe-citrate complex simultaneously, as in experiments with injections of B‑DNICs with NAC or GSH. The signal intensity corresponded to the M-DNIC concentration, which was no more than 1–2 μmol/kg wet tissue (Fig. 3b). The subsequent 30-min incubation of the samples in 200 mM DETC solution gave rise to MNIC−DETC at concentrations 10–5 μmol/kg wet tissue (Fig. 3d). When these samples with MNIC−DETC were then treated with dithionite, the MNIC−DETC level remained practically unchanged (Fig. 3f). In contrast to experiments with administration of ready-made B‑DNICs followed by liver treatment with DETC in vivo or in vitro, the levels of MNIC−DETC forming in vitro were much lower than in vivo (Fig. 2). Similar experiments with GS−NO administration to mice revealed no significant differences in the levels of MNIC−DETC in liver samples.

In experiments with GS−NO administration without iron citrate complexes, no M-DNIC was detected in the liver samples, and the level of MNIC−DETC that emerged in the liver after treatment with DETC did not exceed 0.15 μmol/kg.

MNIC−DETC formation in liver tissue of animals treated in vivo with sodium nitrite or isosorbide dinitrate. Sodium nitrate combined with the iron citrate complex was administered to mice, and after 30 min liver samples were treated with DETC in vivo or in vitro: by additional injection of 500 mg/kg (3 mM/kg) of DETC and sacrifice after 30 min or by 30-min incubation of liver samples in 200 mM DETC, respectively. In both cases, MNIC−DETC formed in the liver at approximately equal concentrations, 20 ± 5 μmol/kg wet tissue, as assessed from the characteristic EPR signal (Fig. 4a). In experiments without DETC, nitrite induced only the formation of nitrosyl hemoglobin complexes, as seen from EPR spectra of liver samples (Fig. 4a). Nitrite injections did not result in the formation of M-DNIC in liver samples from animals that received B-DNIC or GS−NO, judging from the absence of the 2.03 signal from the EPR spectra (Fig. 4a). This result points to the inability of single nitrite injections to induce M-DNIC in animal tissues. It was repeatedly observed in our studies of DNICs in living organisms over the course of more than 5 decades [5, 6, 12–16].

Fig. 4.
figure 4

EPR spectra recorded in liver samples from mice: (a, b) 30 min after intraperitoneal injection of sodium nitrite or ISDN; (c, d) with subsequent intraperitoneal DETC injection 30 min before sacrificing; (e) with subsequent treatment of the samples with DETC and then (f) dithionite.

Full size image
Fig. 5.
figure 5

A schematic presentation of the formation of M-DNICs with thiol-containing ligands induced with the presence of Fe2+ and thiols by (a) S-nitrosothiols or (b) gaseous NO [5, 6].

Full size image

Another significant difference noted in experiments with nitrite injections, on the one hand, and injections of B-DNIC or GS−NO, and incubation of mouse liver with gaseous NO, on the other hand is a threefold rise of the MNIC−DETC level in mouse liver samples with nitrite after their treatment with the potent reducing agent dithionite (Fig. 4e).

The threefold increase in MNIC−DETC after dithionite treatment was also observed in experiments with livers of mice that received the water-soluble form of nitroglycerin, ISDN, almost at the same dose as nitrite (Figs. 4b, 4d, and 4f). This similarity to liver samples from mice that received nitrite might be caused by the fact that ISDN metabolism in the body produced not only NO but also nitrite anions, as indicated in the description of Isoket® [17].

It should be noted that the liver levels of MNIC−DETC in experiments with nitrite and ISDN administered without the iron citrate complex were three to fourfold lower than the reported values (data not shown).

DISCUSSION

Prior to discussing these results, we consider possible pathways of the formation of paramagnetic MNIC−DETC that could act in our experiments on mouse treatment with B-DNIC−GSH/NAC, GSNO, nitrite, or ISDN and in treatment of isolated liver tissue samples with gaseous NO. Numerous studies of this process point to two mechanisms by which MNICs with dithiocarbamate derivatives (DCDs), e.g., DETC or MGD, may form [18–27]. First, with the presence of DNICs in animal tissues, DCDs can take mononitrosyl iron fragments from already present dinitrosyl iron groups to form paramagnetic EPR-active MNICs of the formulas (DCD)2–Fe2+–NO or (DCD)2–Fe+–NO+, as shown in Fig. 1 [18, 20, 23, 25, 26]. Second, without DNICs but with NO in the tissue, the latter can equally efficiently bind to Fe2+ or Fe3+ ions to form, respectively, paramagnetic [(DCD)2–Fe2+–NO] or diamagnetic EPR-silent [(DCD)2–Fe3+–NO] complexes [19, 21–24, 26, 27]. The latter can readily be reduced to the paramagnetic form by various agents, e.g., dithionite, cysteine, or ascorbate [19, 21, 23–27]. As the majority of iron ions (loosely bound, incorporated in MNICs with DCD) in animal tissues in vivo have the oxidation degree 3 [26], just the dramatic threefold increase in the MNIC−DCD level with the presence of various reducing agents should distinguish the second pathway from the first.

This difference in responses to dithionite treatment of isolated liver tissue was observed between samples from mice treated in vivo with sodium nitrite or ISDN, on the one hand, and from mice treated in vivo with B-DNIC−GSH/NAC or GS−NO, as well as after treatment of liver samples with gaseous NO in vitro, on the other hand. This observation allows us to state that the dominating agent is NO in the former case and B-DNIC in the latter. In fact, if B-DNIC−GSH/NAC or GS−NO would entirely decompose in the liver within 30 min and MNIC−DETC would be formed by NO released from the injected compounds and bound to DETC or by the involvement of nitrite arising from NO oxidation, then dithionite, a strong reducing agent, would raise the level of paramagnetic EPR-active MNIC−DETC. The complete absence of this phenomenon from all liver samples from mice treated in vivo with B-DNIC−GSH/NAC or GS−NO, as well as from samples treated in vitro with gaseous NO makes it likely that DNICs with thiol-containing ligands formed in mouse liver in vivo in all these cases. The formation of their mononuclear form is evident from the 2.03 signal (Figs. 2a, 2b, 3a, and 3b), and the presence of B-DNICs at concentrations severalfold higher than M-DNICs is apparent from the proportionally higher integrated intensity of the MNIC−DETC signal compared to the 2.03 signal.

The threefold increase in MNIC−DETC level after dithionite treatment in liver samples from mice treated in vivo with nitrite or ISDN (Figs. 4e and 4f) points to the presence of two MNIC−DETC species. One of them is paramagnetic, EPR-active, and its formula is (DETC)2–Fe2+–NO. The other is diamagnetic, EPR-silent, with the formula (DETC)2–Fe3+–NO. The interaction of the latter with dithionite is responsible for the threefold increase in the level of EPR-active MNIC−DETC. For MNIC−DETC, which is EPR-active without dithionite and which constituted approximately 30% of the total amount of MNIC−DETC, it could be produced in the reaction between DETC and B-DNIC that might have formed in the liver tissue contacting with nitrite.

This assumption is not in line with comparisons of the hypotensive effects of B-DNIC−GSH and sodium nitrite. It has been shown that in terms of doses the hypotensive action of nitrite is more than 100-times weaker than B-DNIC [28]. This fact infers that nitrite injected to an animal body cannot efficiently produce DNICs with thiol-containing ligands; hence, by no means are 30% of NO molecules produced by nitrite could be involved in DNIC formation. Obviously, these molecules bound to the Fe2+–DETC complex to ensure the formation of originally paramagnetic EPR-active MNIC−DETC, not reacting with dithionite.

The question arises as to how GS−NO could ensure the formation of both M- and B-DNICs in the liver in vivo in spite of being administered at the threefold lesser dose in our experiments. The level of EPR-active MNIC−DETC detected in liver samples in the reaction of DNIC and DETC (10–15 μmol/kg tissue) was approximately the same as the level of the same MNIC arising in the livers of mice treated with nitrite in vivo in the interaction between the formed NO and the Fe2+-DETC complex (20 ± 5 μmol/kg tissue). In other words, why was NO sufficient for DNIC formation in mice that received GS−NO, while the same level of nitrite was not?

The different effects of GS−NO and nitrite on DNIC formation in the animal body may be explained by the pathways we proposed for the synthesis of these complexes, assuming the involvement of S-nitrosothiols (RS−NO) or free NO molecules, as shown in Fig. 5.

A significant difference between these pathways is that the formation of one dinitrosyl iron group in a DNIC from RS−NO and free NO requires two and three molecules of these agents, respectively. In both cases, the process is triggered by disproportionation (reciprocal one-electron redox reaction) of two molecules of RS−NO and NO. It is sufficient for the formation of a dinitrosyl iron group with RS−NO, whereas the process with NO demands a third molecule to enter the iron coordination sphere and replace the released nitroxyl molecule HNO [5, 6, 9, 29–32]. It is obvious that in this case the feasibility of the formation of a dinitrosyl iron group determined by the three-step mode of the process (first, binding of two NO molecules to an iron atom; second, release of nitroxyl from the coordination sphere of the iron atom; and third, incorporation of the third NO molecule) should delay the formation of the said group considerably as compared to the time of its synthesis with only two GS−NO molecules. Moreover, with regard to the putative instability of intermediates, this may hamper the formation of dinitrosyl iron groups in liver tissue from NO molecules, whose level in the tissue is relatively low (approximately 0.1 mM), as in our experiments. It is clear that only a significantly higher NO level in the tissue, as with the treatment of isolated mouse liver samples with gaseous NO, could allow the formation of M- and B-DNICs with thiol-containing ligands (Figs. 3a, 3c, and 3e). In the corresponding experiment, liver samples were placed to 100-mL Thunberg vials and exposed to gaseous NO at pressures 100–150 mmHg, the total level of NO being 1 mmol. As the volume of one liver sample was 1 cm3, this level was sufficient to produce up to 0.3 mol of iron dinitrosyl groups per 1 kg of wet tissue.

In discussing this issue, the close relationship between RS−NO and DNIC on the one hand and thiol-containing ligands on the other hand should be taken into consideration. As shown in [5, 6, 30], it is conceivable that these two interconvertible compound types form a self-sustaining and self-regulating chemical system, whose oscillatory regime generates primarily either DNIC or RS−NO.

Previously, we suggested that the formation of B‑DNICs with thiol-containing ligands involved NO produced by iNOS in the mouse liver [7. 8]. We induced iNOS production by treating mice with bacterial lipopolysaccharide as a potent proinflammatory agent. The emergence of B-DNIC in mouse liver was judged from the detection of MNIC−DETC. For this purpose, Fe2+–citrate complex was administered to mice 30 min before sacrificing, and the liver tissue was treated with DETC in vitro. It was supposed that the added iron would bind endogenous NO and form B-DNIC with endogenous thiols. In fact, B-DNICs incorporated up to 70% of the NO produced by iNOS. The total NO level produced by iNOS was assessed from the level of MNIC−DETC formed in the liver in vivo after the injection of iron complexes with DETC as selective NO traps.

Later, we found that the level of MNIC−DETC detected in those experiments increased three to fourfold after treatment with dithionite, as was in experiments with administration of nitrite or ISDN to mice [26]. Thus, it seemed conceivable that endogenous NO occurring in liver samples was be incorporated not to B-DNIC but directly to MNIC−DETC, part of which, with the formula (DETC)2–Fe2+–NO, was recorded by EPR without dithionite treatment, and the other part, EPR-silent (DETC)2–Fe3+–NO, became EPR-active after one-electron reduction by dithionite.

This scenario is ruled out by the ample presently available data that immunocompetent cells, such as macrophages, activated by, e.g., bacterial lipopolysaccharide, can produce considerable amounts of M‑DNICs with thiol-containing ligands [33–39], which come from the activated macrophages to their target cells [36]. As shown in [39], M-DNICs produced in activated macrophages incorporate all their loosely bound (soluble) iron (this iron is incorporated into DNICs.) Therefore, it is hardly likely that these cells also contain binuclear forms of the complexes. In this respect, isolated cells that produce DNICs differ significantly from cells constituting animal tissues and organs. Our studies show that DNICs with thiol-containing ligands are present in tissues and organs in vivo mainly in the binuclear form. Thus, it is reasonable to suggest that the same form arises in the immune system of mice activated with lipopolysaccharide, because this form is responsible for the appearance of paramagnetic MNIC–DETC, EPR-active irrespective of dithionite addition, in liver samples. The other MNIC−DETC species, (DETC)2–Fe3+–NO, which became EPR-active with dithionite, may form with the participation of nitrite, emerging as a result of oxidation of NO produced by iNOS and this species may constitute the majority of MNIC−DETC in liver samples from mice treated with lipopolysaccharide.

The stability and in vitro conversions of NO, nitrite, GS−NO, and B-DNIC−GSH added to homogenized human, ovine, and murine placentae were studied recently [40]. This work is in line with our above-presented studies that concern the conversion of the same agents (except for NO) in vitro in mouse liver samples and in vivo metabolism in mice within 30 min prior to sacrificing. The comparison of results presented in [40] and obtained by our team reveals both similarities and differences. Both works demonstrated M- and B-DNIC−GSH degradation in vitro. In [40], this degradation was evident from the decrease in the levels of these complexes assessed from the amount of gaseous NO released from DNICs in their decomposition with the presence of ferricyanide. In our work, this degradation was apparent from the fact that the amount of MNIC−DETC formed in vitro in liver samples with DNICs decreased notably after sample incubation in DETC solutions in comparison to the amount of MNIC−DETC formed in vivo in mouse liver after intraperitoneal injections of first B-DNIC−NAC/GSH and then DETC (Figs. 2c, 2e and 2d, 2f, respectively). On the grounds of the results obtained in [40] and in our experiments conducted in 1960–1980 [5, 6], it is reasonable to state that the decrease in MNIC−DETC level detected in our work was irrelevant to DETC; rather, it was caused by direct degradation of DNICs with thiol-containing ligands formed in liver in vivo as a result of B-DNIC−NAC/GSH injection followed by ex vivo incubation at room temperature. This degradation might be caused by acidosis developing in the tissue under hypoxic conditions. Acidosis, that is, acidification of the intracellular milieu, could cause protonation of thiol groups in the thiol ligands in DNICs, resulting in DNIC decomposition.

If so, why did we observe the significant increase in the MNIC−DETC level in liver samples exposed to gaseous NO and then kept in DETC solution (Figs. 3a and 3c)? There is no unambiguous answer so far. It was proposed in [40] that animal tissues possess enzyme systems that are inducible under anoxic conditions that degrade DNICs. Another, simpler cause of the phenomenon is also conceivable: alkalization of the intracellular milieu by gaseous NO in DNIC synthesis in liver tissue, which neutralizes acidosis. It follows from the pathway of DNIC formation under these conditions illustrated in Fig. 5 that the emergence of nitroxyl anions (NO–) in the hypothesized disproportionation of NO as the key step in DNIC synthesis should result in milieu alkalization owing to proton scavenging by nitroxyl anions followed by irreversible conversion of nitroxyl molecules HNO to nitrous oxide and water. Similarly, the alkalization of intracellular milieu might occur in DNIC synthesis from the appropriate RS−NOs. According to the formation pathway of these complexes presented in Fig. 5, the protonation of the thiol anion released from RS−NO could be responsible for alkalization of the intracellular environment of DNICs.

The results of both our study and [40] indicate that one of the RS−NO family, GS−NO, can pass into a DNIC. Obviously, this process involves loosely bound iron and exogenous or endogenous thiols. The emergence of DNICs in our studies and, correspondingly, MNIC−DETC in the liver of mice that received GS−NO demanded exogenous iron, is noteworthy. In experiments with nitrite [40], its addition to human placenta homogenate resulted in a notable degree of its conversion to DNIC. Our experiments did not reveal such conversion in animals after injection of a significant nitrite dose, 0.7 mmol/kg animal weight. It is likely that MNIC−DETC was produced in liver samples from these animals by the binding of NO formed from nitrite to Fe2+/Fe3+–DETC complexes rather than in the reaction between DETC and DNICs.

Still, as shown in our earlier study [8], in in vitro experiments where DETC was replaced with glutathione, cysteine, or NAC, significant amounts of DNICs with the said thiol-containing ligands emerged in liver samples from mice treated with nitrite in vivo. This effect was most pronounced when DETC was replaced by NAC, a thiol efficiently penetrating through cell membranes. Similar results were obtained in our studies of 1970–1980s, where significant amounts of M-DNICs with thiol-containing ligands formed in vitro in liver tissue and its homogenate contacting with 0.5–1.0 mM sodium nitrite [5, 6]. In both approaches, notable DNIC levels formed as a result of increase in the levels of their components: thiols or NO, respectively.

CONCLUSIONS

Our data show that when DNICs with thiol-containing ligands are present in animal tissues in vivo, having been either administered to the body or produced with the participation of endogenous NO, they exist mainly in the EPR-silent diamagnetic form. This can be detected by injecting a dithiocarbamate derivative (e.g., DETC) or treating isolated organs and tissues with water solutions of the same agent. This procedure converts diamagnetic DNICs to EPR-active paramagnetic MNICs with the corresponding dithiocarbamate ligands.