Ghost protein damage by peroxynitrite and its protection by melatonin

We have studied the effect of peroxynitrite (ONOO-) on the membrane cytoskeleton of red blood cells and its protection by melatonin. Analysis of the protein fraction of the preparation by SDS-PAGE revealed a dose-dependent (0-600 μM ONOO-) disappearance at pH 7.4 of the main proteins: spectrin, band 3, and actin, with the concomitant formation of high-molecular weight aggregates resistant to reduction by ß-mercaptoethanol (2%) at room temperature for 20 min. These aggregates were not solubilized by 8 M urea. Incubation of the membrane cytoskeleton with ONOOwas characterized by a marked depletion of free sulfhydryl groups (50% at 250 μM ONOO-). However, a lack of effect of ß-mercaptoethanol suggests that, under our conditions, aggregate formation is not mediated only by sulfhydryl oxidation. The lack of a protective effect of the metal chelator diethylenetriaminepentaacetic acid confirmed that ONOO--induced oxidative damage does not occur only by a transition metal-dependent mechanism. However, we demonstrated a strong protection against cytoskeletal alterations by desferrioxamine, which has been described as a direct scavenger of the protonated form of peroxynitrite. Desferrioxamine (0.5 mM) also inhibited the loss of tryptophan fluorescence observed when the ghosts were treated with ONOO-. Glutathione, cysteine, and Trolox® (1 mM), but not mannitol (100 mM), were able to protect the proteins against the effect of ONOOin a dose-dependent manner. Melatonin (0-1 mM) was especially efficient in reducing the loss of spectrin proteins when treated with ONOO(90% at 500 μM melatonin). Our findings show that the cytoskeleton, and in particular spectrin, is a sensitive target for ONOO-. Specific antioxidants can protect against such alterations, which could seriously impair cell dynamics and generate morphological changes. Correspondence


Introduction
Peroxynitrite anion (ONOO -), a highly reactive and biologically important species, is produced under physiological conditions and in vivo by the reaction of superoxide anion radical (O 2 • -) with nitric oxide ( • NO), (1).Nitric oxide, identified as endothelium-derived relaxing factor, is formed during the conversion of Larginine to L-citrulline by an NO-synthase.Endothelial cells, macrophages, neutrophils and neuronal cells have been shown to produce • NO (2,3).
At physiological pH, ONOO -can be protonated and decomposed to an intermediate with reactivity similar to that of hydroxyl radical and nitrogen dioxide (1,(4)(5)(6).Peroxy-P.Di Mascio et al. nitrite can also react directly with sulfhydryl groups (7).The lifetime of this anion is sufficient for diffusion through the membrane and interaction with hydrogen peroxide (8) and cellular constituents (9).Peroxynitrite oxidizes low molecular weight sulfhydryls to disulfides that can be partially recovered by the glutathione-glutathione reductase system but leads to higher sulfur oxidation states of protein sulfhydryls (7).This damage also occurs in the absence of transition metals (1,6,9).The peroxynitrite anion reacts rapidly with carbon dioxide at a rate constant of 5.8 x 10 4 M/s at 37 o C, producing an adduct whose structure is proposed to be nitrosoperoxocarboxylate (ONOOCO 2

-
) and may modulate various biological peroxynitrite-mediated processes (10)(11)(12).Peroxynitrites are able to react with lipids, DNA, proteins and small antioxidant molecules such as glutathione (13,14).These rapid and specific reactions are likely to inactivate important cellular targets.It has been shown that ONOO -promotes lipoperoxidation, protein nitration and a decrease of intracellular reduced glutathione in human erythrocyte (15).
Red blood cell (RBC) shape and deformability are regulated by a submembrane cytoskeleton whose major proteins responsible for regulating the membrane topography are: spectrin, actin, ankyrin, band 4.1, band 4.9, and tropomyosin.These proteins are arranged in a network connected to integral membrane proteins along the bilayer membrane through associations between ankyrin and band 3 and between band 4.1 and glycophorin (16).Structural alterations in these membrane proteins can lead to a loss of deformability, which is essential for RBC passage into small blood vessels of specific organs and tissues.
Red blood cell membrane ghosts offer a good model for studying protein damage induced by ONOO -because their protein composition is well known and they lack organelles, which makes them a simple and suitable biological system.

Synthesis of peroxynitrite
Peroxynitrite was synthesized in a quenched flow reactor as described by Beckman et al. (1).Excess H 2 O 2 was removed by passage over MnO 2 .The solution was then filtered twice and frozen at -20 o C for two days.Peroxynitrite forms a dark yellow top layer due to freeze fractionation, which was removed and stored at -80 o C for further experiments (17).This solution contained 200-400 mM ONOO -as determined by absorbance at 302 nm (e 302 = 1,670 M/ cm).

Preparation of red blood cell ghosts and SDS-PAGE
Mouse blood was freshly collected into ACD (75 mM trisodium citrate, 42 mM citric acid and 139 mM glucose).After washing with phosphate-buffered saline (PBS), the leukocytes and platelets were removed by cellulose filtration (18).The eluate from the column was washed three times with PBS.Erythrocyte ghosts were prepared according to the method described by Tyler et al. (19).RBC lysis was performed by the addition of 20 volumes of lysis buffer (5 mM phosphate, pH 8.0, and 0.1 mM ethylenediaminetetraacetate (EDTA)).After lysis, the ghosts were sedimented at 19,000 g for 10 min and the pellet was washed in lysis buffer until hemoglobin was eliminated.The final pellet was resuspended in PBS.

Exposure of the membrane cytoskeleton to peroxynitrite
Peroxynitrite concentration was measured by the increase in absorbance at 302 nm in 1.2 M NaOH.The membrane cytoskeleton (~200 µg proteins/ml) was incubated with synthesized ONOO -in 50 mM potassium phosphate buffer, pH 7.4.The reaction was performed by placing a small aliquot of ONOO -(2-6 µl) on the side of a tube containing the membrane cytoskeleton solution immediately followed by vigorous vortexing.
As a control for the potential effects of nitrite, nitrate and H 2 O 2 , traces of which can be present in synthesized peroxynitrite, ONOO - was allowed to decompose for 5 min in potassium phosphate buffer, pH 7.4, before the addition of the cytoskeleton.To insure that ONOO -was decomposed, we measured the absorbance of ONOO -in 50 mM potassium phosphate buffer, pH 7.4, at 302 nm.After 5 min, the absorbance at 302 nm of the decomposed ONOO -was the same as that of the buffer alone.The membrane cytoskeleton was incubated in the presence of peroxynitrite for 20 min at room temperature.After incubation, erythrocyte ghosts were sedimented, the degradation products contained in the buffer were eliminated, and the final pellet was dissolved in 8 M urea/1% SDS for protein and thiol measurement or in electrophoresis sample buffer containing 10% SDS and 2% mercaptoethanol.

SDS-polyacrylamide gel electrophoresis
The samples were solubilized in electrophoresis sample buffer and submitted to SDS-PAGE in a 3.5-12% polyacrylamide gradient gel (20).The gels were stained with Coomassie brilliant blue R-250 and the protein bands were scanned with a Schimadzu (C-9000, Japan) densitometer.

Protein measurement
Membrane cytoskeleton protein was measured by the method of Lowry et al. (21) and/or according to the method of Bradford (22) (Biorad assay; Biorad Laboratories, Hercules, CA, USA).

Protein sulfhydryl group determination
Sulfhydryl groups were measured using the DTNB procedure described by Di Monte et al. (23).

Fluorescence measurement
Oxidation of tryptophan (excitation l = 295 nm and emission l = 310-450 nm) or melatonin (excitation l = 272 nm and emission l = 290-450 nm) was monitored by changes in fluorescence.Fluorescence spectra were recorded on a Spex 1681 spectrometer (Spex, Edison, NJ, USA) coupled to a 386DX personal computer.

Results and Discussion
Exposure of RBC ghost membrane cytoskeletons to ONOO -for 20 min induced modifications in the polypeptide pattern observed by SDS-PAGE (Figure 1A and B).In the peroxynitrite-exposed cytoskeleton, two major changes in the polypeptide composition of the ghost proteins were evident, namely the appearance of a large-molecular weight aggregate (>200 kDa) and a dosedependent decrease in the amount of polypeptides migrating in the gel, with a major effect on spectrin bands.Two bands between spectrin and band 3 with a molecular weight <198 kDa were observed, which dis-  all polypeptides (Figure 1A, Band 6).The disappearance of these proteins is consistent with an increase in the size of the pellet obtained prior to electrophoresis (data not shown).Loss of polypeptides through a hydrolytic mechanism of reactive oxygen species as described by Davies and Delsignore (24), and particularly of ONOO -(25) can be excluded by measuring the total amount of ghost proteins and comparing this with the solubilized pellet content.No difference in the quantity of proteins was detected in the control or treated samples.Mallozzi et al. (26) investigated the effects of ONOO -on human erythrocytes and obtained two different responses.At low concentration (<100 µM) ONOO -stimulated a metabolic response of phosphorylation activity and at high concentration (>200 µM) it caused cross-linking of membrane proteins, nitration of tyrosines and massive methemoglobin production.
Among the major targets of ONOO -are R-SH groups (7). Figure 2 shows a dosedependent decrease of R-SH groups (50% at 250 µM ONOO -) in ghost proteins when treated with ONOO -.The extent of lesion to the spectrin bands was slightly different when ONOO -treatment was performed in the presence of urea or under reducing conditions using ß-mercaptoethanol.This result suggests that crosslinking reactions other than disulfide formation contribute to aggregate formation.Thiol groups and tryptophan have been shown to be susceptible to ONOO - attack (7).Similarly, tryptophan residues from the ghost proteins are affected by ONOO -.The loss of R-SH shown in Figure 2 is paralleled by a decrease in fluorescence intensity at 330 nm after excitation at 295 nm after treatment with different ONOO -concentrations.Analysis of the fluorescence spectra after treatment in the presence of 0.5 mM desferrioxamine (Figure 3, b) or 0.5 mM DTPA (Figure 3, c) showed that desferrioxamine has a protecting effect compared to 0.5 mM ONOO -alone.The loss of tryp- tophan fluorescence induced by ONOO -is due to the formation of nitrotryptophan, visible with the presence of a yellow color (27).These experiments were confirmed also using tryptophan (data not shown).Desferrioxamine (0.5 mM) is able to protect against the disappearance of the spectrin bands (Figure 4).This confirms the results obtained for tryptophan fluorescence with 0.5 mM desferrioxamine (Figure 3), which can be explained by the scavenging effect of desferrioxamine on trans-peroxynitrite (28).Diethylenetriaminepentaacetic acid (0.5 mM) was a weak inhibitor of the ONOO -reaction (Figure 4), chelating transition metals involved in the process (1,10), but otherwise having no direct effect.Mannitol at 100 mM was used in the presence of ONOO -to evaluate the contribution of hydroxyl radicals to ghost damage.Only a 10% protective effect on the disappearance of spectrin bands could be observed (data not shown).This result shows that hydroxyl radicals are not the main reactive species in this system (11,12,28,29).However, nitrogen dioxide formed by the decomposition of ONOO - may also be involved in this damage (30).Glutathione and cysteine (1 mM) were highly effective in protecting spectrin against ONOO --mediated damage (Figure 5).It is known that glutathione is very important for the maintenance of erythrocyte integrity.Similarly, plasmid DNA damage caused by ONOO -was inhibited by these compounds in vitro (12,14).
Melatonin has been demonstrated to be a very efficient scavenger of reactive oxygen species, especially hydroxyl radicals, in biological systems (31) cinogen safrol (34,35).The use of this hormone in our model revealed that this is a very good protector of erythrocyte proteins against ONOO --induced damage.In fact, with 500 µM melatonin, the protective effect was almost 90% (Figure 6).At lower concentrations, melatonin was more efficient than glutathione or cysteine.At physiological concentrations (50-1000 pg/ml plasma) (36), melatonin was also able to control the quantity of • NO generated via the NO-synthase reaction (37).Since melatonin has an indole moiety, like tryptophan, we also measured the fluorescence spectra of melatonin in the presence of different ONOO -concentrations (Figure 6, inset).As in the tryptophan experiments, a loss of fluorescence intensity was observed, showing that melatonin scavenges peroxynitrite.
Trolox ® (1 mM), the more water-soluble a-tocopherol analog, was also able to protect against protein damage, in a manner similar to melatonin (data not shown).Peroxynitrite-induced erythrocyte hemolysis was also demonstrated to be partially suppressed by Trolox ® and other antioxidants (38).Interestingly, another class of antioxidants, the pyrrolopyrimidines, prevented nitrotyrosine formation in peroxynitrite-treated red blood cell membranes, but had little effect on membrane cross-linking (39).The present data show that the erythrocyte cytoskeleton is a sensitive target for ONOO -.Spectrin is the main target of ONOO -, and can be protected by melatonin and thiol antioxidants.Such alterations of spectrin could seriously impair cell dynamics and generate morphological changes, and can be prevented by specific antioxidants.

P
. Di Mascio et al.

Figure 2 -
Figure 2 -Measurement of sulfhydryl groups (open circles) and tryptophan fluorescence intensity (filled circles) after reaction with different concentrations of peroxynitrite for 20 min at room temperature.

Figure 6 -
Figure 6 -Red blood cell ghosts cytoskeleton protection by melatonin in the presence of peroxynitrite.Inset: Fluorescence spectra of 10 µM melatonin when treated with ONOO -: a) no ONOO -, b) 0.5 mM ONOO -, c) 1 mM ONOO -and d) 2.5 mM ONOO -, for 20 min at room temperature.