- Email: [email protected]
Stable Nitroxide Radicals Protect Lipid Acyl Chains From Radiation Damage
Free Radical Biology & Medicine, Vol. 22, No. 7, pp. 1165–1174, 1997 Copyright q 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/97 $17.00 / .00
PII S0891-5849(96)00509-6
Original Contribution STABLE NITROXIDE RADICALS PROTECT LIPID ACYL CHAINS FROM RADIATION DAMAGE Ayelet M. Samuni and Yechezkel Barenholz Department of Biochemistry, Hebrew University, Hadassah Medical School, Jerusalem 91120, Israel (Received 6 April 1996; Revised 12 August 1996; Accepted 23 September 1996)
Abstract—The present study focused on protective activity of two six-membered-ring nitroxide radicals, 2,2,6,6tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy-Tempo (Tempol), against radiation damage to acyl chain residues of egg phosphatidylcholine (EPC) of small unilamellar vesicles (SUV). SUV were g -irradiated (10–12 kGy) under air at ambient temperature in the absence and presence of nitroxides. Acyl chain composition of the phospholipids before and after irradiation was determined by gas chromatography. Both Tempo and Tempol effectively and similarly protected the acyl chains of EPC SUV, including the highly sensitive polyunsaturated acyl chains, C20:4, C22:5, and C22:6. The conclusions of the study are: (a) The higher the degree of unsaturation in the acyl chain, the greater is the degradation caused by irradiation. (b) The fully saturated fatty acids palmitic acid (C16) and stearic acid (C18) showed no significant change in their levels. (c) Both Tempo and Tempol provided similar protection to acyl chain residues. (d) Nitroxides’ lipid-bilayer/aqueous distribution is not validly represented by their n-octanol/ saline partition coefficient. (e) The lipid-bilayer/aqueous partition coefficient of Tempo and Tempol cannot be correlated with their protective effect. (f) The nitroxides appear to protect via a catalytic mode. Unlike common antioxidants, such as a-tocopherol, which are consumed under irradiation and are, therefore, less effective against high radiation dose, nitroxide radicals are restored and terminate radical chain reactions in a catalytic manner. Furthermore, nitroxides neither yield secondary radicals upon their reaction with radicals nor act as prooxidants. Not only are nitroxides self-replenished, but also their reduction products are effective antioxidants. Therefore, the use of nitroxides offers a powerful strategy to protect liposomes, membranes, and other lipid-based assemblies from radiation damage. q 1997 Elsevier Science Inc. Keywords—Spin labels, Lipid peroxidation, Liposomes, Antioxidants, SOD-mimic, Superoxide
carriers for a broad spectrum of agents in medical and nonmedical applications.2 – 4 Phosphatidylcholines (lecithins) and cholesterol are the major components of eukaryotic biological membranes, and of many lipid-based pharmaceutical formulations. Lecithins are a heterogeneous group of compounds whose constituent fatty acids vary in chain length and degree of unsaturation. Most naturally occurring lecithins (such as in egg yolk) contain substantial amounts of polyunsaturated fatty acids (PUFA), which are particularly susceptible to oxidation mediated by free radicals.5 – 8 The protection of phospholipid-based formulations is better controllable than that of membranes in vivo. Preventive and protective measures are generally taken to minimize oxidative damage of lipid-based emulsions
INTRODUCTION
Lipid oxidation, also called lipid peroxidation (LPO), involves various processes that lead to the formation of a broad spectrum of products and intermediates, including peroxides, alkoxides, aldehydes, and fatty acids. It has long been known as a problem in the preparation and preservation of lipid matrix and liposomal dispersions and formulations, and has been implicated in degeneration processes in aging and various pathological conditions.1 LPO is also a major obstacle to expanding the use of formulations that are based on lipid assemblies, such as emulsions and liposomes, as Address correspondence to: Ayelet M. Samuni, Department of Biochemistry, Hebrew University, Hadassah Medical School, Jerusalem 91120, Israel. 1165
/ 2b27 2426 Mp 1165 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1166
A. M. SAMUNI and Y. BARENHOLZ
and liposomal formulations. In addition to using the proper conditions of temperature, pH, and light, to minimize LPO, the traditional means of intervention include the use of chelating agents, such as Desferal, which neutralize redox-active metals, and lipophilic antioxidants, such as a-tocopherol (vitamin E), which act as a scavenger of organic free radicals.9,10 Many antioxidants are approved for use in food products and some have been used to preserve liposomes. No systematic studies to establish the ideal antioxidant have been carried out on liposome systems. The literature contains many disagreements as a result of differences in experimental details, which make it difficult to compare the various antioxidants. Unlike isotropic oil systems, mono- and bilayered assemblies like micelles, emulsions, liposomes, and biological membranes are anisotropic structures having a well-defined organization. Consequently, the distribution of antioxidant in the lipid assembly, the location of its active moiety in the anisotropic system and its lateral diffusion rate might be crucial to the antioxidative activity.11 Watersoluble antioxidants like ascorbic acid are reportedly less effective by themselves in preventing lipid peroxidation, although they improve protection by regenerating lipophilic antioxidants.12,13 A newer strategy that has been introduced employs stable nitroxide radicals to control degradation processes mediated by deleterious reactive species. Nitroxides are stable, nontoxic, nonimmunogenic,14 – 17 cyclic radicals, of diverse size, charge, lipophilicity, and permeability through biological membranes, which have been widely used as biophysical probes, spin labels, and contrast agents for nuclear magnetic resonance imaging.18 – 20 Inhibition of LPO in rat liver microsomes by both five- and six-membered-ring nitroxides has been previously reported.21 – 23 The antioxidant effects of both lipophilic and hydrophilic nitroxides, as well as of their respective hydroxylamines, indicate the operation of different mechanisms underlying the prevention of LPO.21 – 23 Other studies showed that nitroxides can protect cells, isolated organs, and whole animals against oxidative stress inflicted by diverse insults.24 Radical-mediated peroxidation processes, as in the case of g-irradiation, yield a variety of products, some of which appear only transiently, and undergo fast changes. This is the case for the fatty acid peroxidation indicators malonyl dialdehyde (MDA) and lipid-hydroperoxide (LOOH). As shown by Lang and Vigo-Pelfrey, the accumulation of MDA and LOOH does not validly reflect fatty acid disappearance. In contrast, a quantitative determination of individual fatty acids (using GC) or cholesterol (using HPLC) enables an accurate and valid evaluation of membrane lipid degra-
dation.8 The present study focused on the damage inflicted to acyl chain residues of egg phosphatidylcholine (EPC) by g-irradiation, and the protection provided by nitroxide radicals. g-Irradiation, selected as the means of inducing damage, is a well defined procedure, allowing good control over the dose administered, onset and termination of reaction, uniform exposure to the insult, the radical species formed, and does not require the addition of components to the test sample. The protection by 2,2,6,6-tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (Tempol)
of acyl chains of small unilamellar vesicles (SUV) prepared from EPC, from radiation, and its relevance to the nitroxide concentration in the lipid bilayer and in the vesicle aqueous phase, were studied.
MATERIALS AND METHODS
Materials Egg phosphatidylcholine (containing 95.6% PC, 3.1% sphingomyelin, 0.5% lyso PC and no other lipids, containing 0.12% DL-a-tocopherol) was purchased from Lipoid KG (Ludwigshafen, Germany); L-a-phosphatidylcholine (egg) (ú99%) from Avanti Polar Lipids (Alabaster, AL, USA); n-octyl alcohol from Riedelde haen AG, Seelze-Hannover, Germany; sterile nonpyrogenic saline from Travenol, Israel; 2,2,6,6-tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy2,2,6,6-tetramethylpiperidine-1-oxyl (Tempol) from Aldrich, Milwaukee, WI; Meth-Prep II kit from Alltech Associates, Deerfield, IL. All other chemicals were of analytical grade. Bidistilled water used was also deionized by millipore column.
Determination of phospholipid concentration Phospholipid concentration of the liposome preparation was determined using a modification of the Bartlett procedure.25
/ 2b27 2426 Mp 1166 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
Particle size determination The distribution of the particle size in the liposomal dispersion was measured by photon correlation spectroscopy using a Coulter Model N4 SD apparatus.25
Preparation of liposomes SUV liposomes were prepared from EPC, 12% in sterile nonpyrogenic water, by high-pressure homogenization as described elsewhere,26 using a two-stage high-pressure homogenizer Gaulin LAB 60 (APV Gaulin, Hilversum, Holland). Vesicle size distribution was unimodal, with mean vesicle size of 25 { 5 nm.
Ionizing radiation Standard steady-state radiochemical technique was employed, using a 60Co g-source at the radiotherapy unit, Hadassah Hospital, Ein-Karem, Jerusalem, with a dose rate of 10 Gyimin01. Samples were prepared in test tubes and irradiated at ambient temperature, without stirring, under air, protected from light. Radiation dose varied between 10 and 12 kGy due to experimental restraints. The exact dose is stated for each specific case.
Determination of acyl chain peroxidation Acyl chain composition of the phospholipids before and after irradiation was determined as described by Barenholz and Amselem25 and summarized briefly as follows: after a Bligh and Dyer extraction,27 the lower (chloroform-rich) phase was transferred to a small glass bottle, and evaporated under N2 to complete dryness. To transmethylate the acyl chains of the phospholipids, 50 ml of toluene and 20 ml of Meth-Prep II were added to the extracted lipid. The samples were left for 30 min at RT, and injected (volume of 2 ml) into a Perkin– Elmer AutoSystem gas chromatography (GC) and Autosampler, using a 6-ft 10% Silar 10C column (Alltech), dry N2 as the carrier gas, and a flame ionization detection. The initial temperature of the run was 1407C for 5 min, then the oven temperature was raised at a rate of 57C/min up to 2407C, and then kept there for 5 min. Methyl esters were identified by comparing the retention times with those of known standards. Palmitic acid (C16), a saturated fatty acid, which has been previously found resistant to peroxidation by g-irradiation, was selected as a reference for the determination of the extent of degradation in other acyl chains of EPC.
1167
Electron paramagnetic resonance (EPR) measurements EPR spectrometry was employed to detect and follow the nitroxide free radicals, using a JES-RE3X ESR spectrometer (JEOL Co., Japan). Samples were drawn by a micropippette into a gas-permeable Teflon capillary of 0.81 mm inner diameter, 0.05 mm wall thickness, and 15 cm length (Zeus Industrial Products, Raritan, NJ). Each capillary was folded twice, inserted into a 2.5 mm i.d. quartz tube open at both ends, and placed in the EPR cavity. EPR spectra were recorded with center field set at 3361 G, 100 kHz modulation frequency, 1 G modulation amplitude, and nonsaturating microwave power. Nitroxides decay in biological systems predominantly through a one-electron reduction yielding the respective cyclic hydroxylamines.20 For determination of the total concentration of nitroxide / hydroxylamine, the hydroxylamine was oxidized by 1 mM ferricyanide or 0.1 M H2O2 at pH 12. 28 Nitroxide lipophilicity, n-octanol-to-saline partition To assess nitroxide lipophilicity, the partition between n-octanol and saline was studied and the n-octanol/saline partition coefficient, Koct/aq, of the nitroxide was determined. n-Octanol and saline (1/1; v/v) were thoroughly mixed and a known amount of nitroxide (Tempo or Tempol) was added. Samples were prepared at various initial concentrations of nitroxide. The mixture was vortexed for 5 min (shaking the samples for 24 h gave the same results) and then left for 1 h until separation of phases. The samples were centrifuged for 10 min (7000 1 g) to facilitate separation of the two phases. An aliquot of each of the two phases was sampled and the intensities of the respective EPR signals were compared. For calibration, both Tempo and Tempol were dissolved in n-octanol and in saline at various concentrations in the range of 0.1–1.0 mM, their EPR signals were determined, and the respective calibration curves were obtained. Partition of nitroxide between lipid bilayer and saline Partitioning of the nitroxides between the liposome lipid bilayer and the aqueous phase was measured using a two-compartment Lucite cuvette specially designed for equilibrium dialysis. Each compartment has a volume of 0.3 ml, and the two compartments were separated by a dialysis membrane of molecular weight cut of 12–14,000 daltons. One compartment contained the liposomal dispersion and the other compartment contained saline. The samples were incubated under con-
/ 2b27 2426 Mp 1167 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1168
A. M. SAMUNI and Y. BARENHOLZ
tinuous shaking, at RT, to allow the dialysismembrane-permeable nitroxide to equilibrate between the two compartments. Following 24 h incubation, samples from the liposome-free and liposome-containing compartments were taken, scanned in the EPR spectrometer, and the respective intensities Cliposome and Caq of the EPR signal of nitroxide in each compartment were compared. Because nitroxide EPR signals in the lipid and aqueous phases are different, and the differences are dependent on the specific nitroxide, the intensities of the EPR signal of Tempo and Tempol at various lipid concentrations were measured, and calibration curves were constructed, by means of which the actual nitroxide concentrations in each compartment were calculated. Sample preparation Samples, prepared in saline pH 4.5, or in 10 mM phosphate buffer pH 7.4, contained 5 mM or 20 mM of SUV phospholipid, with an average particle size of 25 { 5 nm. Various radiation doses were tested, and significant loss of PUFA was induced using á 10 kGy; therefore, the samples were g-irradiated with a dose of 10–12 kGy, under air, at ambient temperature, in the absence and presence of nitroxide. Using EPC without vitamin E gave similar results. In all experiments, the preparations and measurements of samples were carried out under air, at RT, with minimal exposure to light.
mination of degradation in other acyl chains of EPC (Fig. 1). The results show increasing damage with decreasing saturation of the fatty acid, reaching over 80% loss for the most polyunsaturated fatty acid, C22:6 (docosahexanoic). The extent of damage increased with the decrease in phospholipid concentration. Addition of formate (0.1 M), which under air converts iOH, iH, and e0aq into secondary Oi20 radicals, in the absence or the presence of 5–50 mM Cu(II) or Fe(II) did not potentiate the damage. Formate alone even slightly decreased the damage. Protection by nitroxides. In nonirradiated liposomal dispersions no effect of 5 mM nitroxide on liposome size or chemical composition was found. Both nitroxides protected the unsaturated acyl chains in a concentration-dependent manner against radiation damage induced in phosphatidylcholine (Fig. 2). Even the most sensitive polyunsaturated fatty acids such as C20:4, C22:5, and C22:6 were fully protected by 5mM nitroxide, whereas 1 mM nitroxide provided partial protection, preventing about 50% of the damage (Fig. 2).
RESULTS
Acyl chain peroxidation PUFA degradation. In studies of LPO, the accumulation of MDA is generally used to assess the extent of damage. The validity of this widely used procedure has been questioned8,29,30 considering various artefacts associated with this assay:25,29,30 (a) most of the MDA detected is generated during the test itself; (b) added metal ions, H2O2, antioxidants, and chelating agents influence peroxide decomposition during the assay itself; and (c) the type and concentration of acid added also influence the amount of thiobarbituric acid reactive substance formed (TBAR). In the present study, the degradation of the acyl chain residues of the phospholipids following 10 kGy irradiation was measured using GC, which enables an accurate quantification of each of the acyl chains without interference of other oxidation products (such as those of cholesterol). Palmitic acid (C16), a saturated fatty acid, which is resistant to peroxidation by g-irradiation, served as an internal reference for the deter-
Fig. 1. Effect of g-irradiation on composition of fatty acids in EPC SUV. Liposomal dispersion of EPC SUV in saline was g-irradiated with 10 kGy at RT. Fatty acid composition was determined after methyl esterification followed by GC separation. The new levels of fatty acids, related to palmitic acid, which served as internal standard, are presented as a fraction of their original levels. Upper— 5 mM phospholipid; Lower—20 mM phospholipid.
/ 2b27 2426 Mp 1168 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
1169
coefficient values, Koct/aq determined for Tempo and Tempol are listed in Table 1. Nitroxide lipid-bilayer/saline partition
Fig. 2. Tempo inhibits radiation damage induced in PUFA. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of several concentrations of 0.1– 5.0 mM Tempo. Fatty acid composition was determined after methyl esterification followed by GC separation. The residual levels of six unsaturated fatty acids, related to palmitic acid, which served as internal standard, are presented as a fraction of their original levels. Inset: Comparison of the inhibitory effect provided by Tempo and Tempol against radiation damage induced in PUFA. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of 1 mM Tempo or Tempol.
Both nitroxides had a similar protective effect, with no statistically significant difference (using the Mann– Whitney U-test), as will be discussed below. Nitroxides’ partition between octanol/saline and liposomal-bilayer/saline Lipophilicities of the nitroxides were evaluated by determining the coefficient of partition between the lipophilic and aqueous compartments.
Coefficients of nitroxide partition between liposomal lipid bilayer assembly and aqueous phase were determined using a two-compartment cuvette specially designed for equilibrium dialysis. This method excludes the need to separate the lipid bilayer from the aqueous phase in a liposomal dispersion. Following 24 h incubation under continuous shaking at RT, samples from the liposome-free and liposome-containing compartments were taken and scanned in the EPR spectrometer. The nitroxide concentration Cliposome and Caq in the two compartments were calculated from their EPR signals according to their respective calibration curves (in the different media). The experiments were repeated using various lipid concentrations, the values of the ratio Cliposome/Caq were plotted vs. the volume fraction, f, occupied by the lipid bilayer (Fig. 3). The volume fraction f was calculated using partial specific volume (vV ) of 0.984 mlig01 for EPC SUVs.31 The experimental values of Cliposome/Caq increased linearly with f, obeying the expression: Cliposome /C aq Å (1 0 f) / fiClipid /C aq Å (1 0 f) / fiK lipid/aq
from which the partition coefficients, Klipid/aq, were calculated. Lipid-bilayer/saline partition coefficients determined for Tempo and Tempol are listed in Table 1. As can be calculated using the respective Klipid/aq values, the lipid phase in a liposome dispersion of 5 mM EPC contains 10.45% and 1.3% of Tempo and Tempol, respectively. Figure 4 demonstrates the degree of protection provided the acyl chain 20:4 as a function of the nitroxides’ (Tempo and Tempol) concentrations in the aqueous and lipid phases. Nitroxides’ fate upon g-irradiation
Nitroxide octanol/saline partition The partition between octanol and water was determined using EPR spectrometry of the nitroxide concentration in the two phases. Octanol/saline partition
EPR spectra of irradiated and nonirradiated samples containing 5 mM nitroxide were measured and compared in order to determine the residual fraction of nitroxide remaining following irradiation in the absence
Table 1. Partition Coefficients of Tempo and Tempol Between Octanol and Saline, Koct/aq, and Between Lipid-Bilayer and Saline, Klipid/aq Nitroxide
Octanol/Saline Koct/aq
Lipid Bilayer/Saline Klipid/aq
Koct/aq:Klipid/aq
Tempo Tempol
85.15 { 9.49 3.70 { 0.46
31.09 { 0.71 3.50 { 0.41
2.74 1.06
/ 2b27 2426 Mp 1169 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1170
A. M. SAMUNI and Y. BARENHOLZ
Fig. 3. Tempo and Tempol lipid-bilayer to saline partition coefficients. Partition coefficients of Tempo and Tempol between lipid-bilayer and saline were measured using equilibrium dialysis double compartment cell. The concentration of nitroxide in lipid-free (Caq) and lipid-containing (Cliposome) solution in two chambers, separated by a semipermeable membrane, were measured by EPR and compared. The lipid concentration was varied and the ratio Cliposome/Caq was determined. The partition coefficient, Klipid/aq, was calculated considering the partial specific volume of EPC SUV, at various lipid concentrations, using the expression: Cliposome/Caq Å (1 0 f) / fKlipid/aq.
and the presence of 20 mM SUV. The results showed that the residual nitroxide (Tempo or Tempol) was dependent on the irradiation dose and on the absence or presence of the lipid. The results are listed in Table 2. To determine the total level of nitroxide plus its respective hydroxylamine, the sample was incubated for 1 h with 0.1 M H2O2 at pH 12 or with 1 mM ferricyanide. No increase in the EPR signal of Tempol or Tempo was found, indicating that no appreciable Tempol-H or Tempo-H, the respective hydroxylamines, were accumulated upon irradiation. DISCUSSION
A common strategy of protection against LPO employs reducing agents that act as preventive and chain-
Fig. 4. Protection of acyl chain C20:4 (arachidonic acid) from radiation as a function of the nitroxide concentrations in the aqueous and lipid phases. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of several concentrations of nitroxide. The extent (%) of protection provided by Tempo (squares) and Tempol (circles) to the acyl chain is plotted as a function of their concentration in the two phases: aqueous phase (open symbols) and lipid bilayer (solid symbols).
breaking antioxidants. However, their efficacy is limited because they (a) are being depleted; (b) give rise to secondary radicals which may be deleterious themselves; and (c) may act as prooxidants.32 Nitroxides, on the other hand, act catalytically and are self-replenished, without any prooxidative effect. They can protect cells and the whole animal from radiation33–36 and represent a new class of nonthiol aerobic radioprotectors. The reactions of nitroxides, being radicals themselves, as antioxidants and radioprotectors do not yield secondary radicals, but rather terminate the radical chain reactions. Nitroxides have been shown to react with peroxyl radicals and inhibit lipid peroxidation,21,22 and their activity has been attributed to their respective hydroxylamines.
Nitroxides inhibit acyl chain degradation Hydrophilic antioxidants were previously reported to be less effective in protection against LPO induced by g-irradiation than lipophilic ones. Their effect was attributed to a regeneration of lipophilic antioxi-
Table 2. Residual Fraction of Nitroxide Remaining Following g-Irradiation in the Absence or Presence of 20 mM SUV 12 kGy
32 kGy
Nitroxide
no SUV
/SUV
no SUV
/SUV
Tempo (5 mM) Tempol (5 mM)
26 { 3 26 { 4
34 { 6 38 { 4
0.4 { 0.1 6.8 { 0.1
17 { 4 17 { 4
/ 2b27 2426 Mp 1170 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
dants.12,13 The rationale behind the use of lipophilic antioxidants is their selective incorporation and retention in the membrane (or lipid-bilayer), thus being in close proximity to the site of damage. The present study evaluates the capacity of the amphipathic nitroxides to protect against g-irradiation-induced LPO (PUFA degradation). We tried to address major issues in optimizing the selection of protecting agents against g-irradiation, especially to assess to what extent their lipophilicity and concentration in the aqueous phase and in the membrane affect their protective effect. We demonstrated that the protective effect is dependent on the concentration of the nitroxide(s) in the aqueous phase, achieving for both Tempo and Tempol complete protection at 5 mM (for 20 mM EPC). As will be discussed below, the protective effect is neither related to the lipophilicity of the nitroxide (Tempo @ Tempol) as assessed by Koct/aq, nor by their concentration in the EPC lipid bilayer (Tempo ú Tempol). The results presented here suggest that for antioxidants that lack prooxidant effect, and that can be regenerated during the oxidation, low concentration in the membrane is not an obstacle to the protection against g-irradiation oxidative damage, as long as sufficient concentration in the aqueous phase can be achieved. The protective effect is concentration dependent, as can be seen in Figs. 2 and 4, achieving complete inhibition of PUFA degradation at 5 mM of Tempo (or Tempol). These results suggest the usefulness of watersoluble antioxidants compared to lipophilic antioxidants, and emphasize the need to further investigate the modes of damage and protection. Deleterious species, protective agents, and mechanism Our results showing no significant effect of formate on the irradiation damage suggest that iOH and Oi20 similarly contribute toward damage of lipid acyl chains during g-irradiation. On the other hand, iOH and Oi20 contribute differently toward nitroxide destruction. While the removal of Oi20 radicals by nitroxide is a catalytic process that does not affect its concentration, the abstraction of H by hydroxyls causes an irreversible destruction of the nitroxide. This explains the lower loss of nitroxide when irradiated in the presence of EPC SUV. The inhibition of nitroxide depletion in the presence of EPC SUV, indicates that its destruction is due to nonspecific reaction damage by the radiolytically formed hydroxyls, rather than its reaction with phospholipid-derived reactive species such as Li, LOi, or LOOi. The antioxidant activity of nitroxides has been previously demonstrated using rat liver microsomes and related largely to their respective cyclic hydroxyl-
1171
amines.21,22 In the presence of active reducing enzymatic systems the nitroxides are rapidly reduced to their respective hydroxylamines, Tempo-H and Tempol-H. In such cases, the protective species might be the hydroxylamine. In the present work the situation is different because the predominant species was found to be solely the nitroxide itself. Also, upon addition of Tempol-H to the test samples, no hydroxylamine was left following irradiation. Instead, Tempol was formed, indicating that Tempol-H was oxidized to Tempol. Both the nitroxide and its hydroxylamine can reduce carbon-centered and oxygen-centered lipid radicals and break the radical chain reaction:
Reactions 1 and 2 yield the nitroxide and its oxidized form, the oxo-ammonium cation, respectively. The latter can be rapidly reduced to its respective nitroxide through a diffusion-controlled reaction with superoxide, having k3 Å 1.5i1010 M01s01, as previously reported.19
Evidently, Oi20 radicals are removed also through reaction with the nitroxide itself:
/ 2b27 2426 Mp 1171 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1172
A. M. SAMUNI and Y. BARENHOLZ
This mechanism accounts for the replenishment of nitroxide with a concurrent removal of Oi20. While protection by nitroxides of lipids against iOH-induced damage is achieved in a suicidal mode that results with their progressive depletion, the mechanism underlying their protective activity against Oi20 induced damage and the breaking of radical chain reaction is catalytic. It is important to compare the nitroxides to traditional antioxidants. Among the lipophilic antioxidants most commonly used to inhibit LPO are tocopherols, and in particular a-tocopherol, a reactions chain breaking agent. Although the mechanism underlying the protective effect of tocopherols in membranes has not been fully elucidated, it is clear that they act in a stoichiometric fashion and are consumed while scavenging and preventing lipid peroxidation.37,38 Obviously, their concentration decreases upon exposure to a stress; therefore, they cannot protect against high doses of g-irradiation. Vitamin E was also shown to reduce transition metal ions, thus enhancing the Fenton reaction and causing a prooxidant rather than antioxidant effect.32 This constitutes another potential drawback in the use of vitamin E as antioxidant. In previous studies, 0.2–1.0 mol% of a-tocopherol (based on phospholipid content) effectively inhibited LPO.13,39,40 Wills reported38 an effective inhibition by 0.005–0.01% vitamin E of LPO in herring oil and starch mixtures irradiated with doses as high as 10 kGy. However, in the present study, 0.12 mol% DL-a-tocopherol in the EPC SUV did not prevent radiationinduced oxidation of EPC PUFA (Fig. 1). Effect of nitroxide lipophilicity and concentration in the membranes The lipophilicity may have a dual effect on the protective activity of an antioxidant against LPO: by (a) dictating its concentration at the site of LPO in the lipid bilayer, as expressed more validly by the bilayer/ aqueous partition coefficient (Klipid/aq) or (b) direct involvement in the antioxidant mechanism of action. The common way to describe lipophilicity is by the oil/aqueous (heptane/aqueous) and n-octanol/aqueous partition coefficient.41 For this reason, many studies on antioxidant activity of nitroxides consider the coefficients determined for partition between water and n-octanol.19,22,42 – 44 Our results show that Tempo’s partitions between n-octanol/saline and lipid-bilayer/saline greatly differ (Çthreefold, see Table 1). Not only does Koct/aq of a given nitroxide differ from its Klipid/aq, but also the ratio Koct/aq/Klipid/aq differs between Tempo and Tempol (Table 1). Koct/aq and Klipid/aq for Tempo and Tempol indicate that the Koct/aq value inaccurately de-
fines the nitroxide concentration in the bilayer. These results are in good agreement with recent studies on other molecules.45,46 This discrepancy is related to the large differences between bulk isotropic systems, such as oil or n-octanol, and the anisotropic microscopic environment of the lipid bilayer.46 This difference remains even when the dielectric constants of the bulk and the bilayer region, in which the antioxidant is localized, are matched (i.e., n-octanol and the bilayer interface region). According to their respective lipid bilayer/saline partition coefficients (3.5 and 31), Tempol and Tempo would have been anticipated to exert different protective effects. In fact, they provided similar protection against radiation-induced LPO. This finding indicates that both the large difference (23-fold) in Tempo and Tempol values of Koct/aq (lipophilicity) and lipid/aq partition coefficient (bilayer concentration), are not correlated with the protective effects of Tempo and Tempol, while their concentration in the aqueous phase is correlated with the antioxidant efficacy (Fig. 4). In previous attempts to determine the partition of nitroxide between aqueous and liposomal compartments, multilamellar vesicle (MLV) dispersions containing nitroxide were centrifuged and the EPR spectra of both the pellet and supernatant were compared.21 However, such an approach is less accurate due to the large interstitial volume of the MLV pellet. Also, small liposomes present in the supernatant may overestimate the level of nitroxide in the aqueous phase. Because variables such as lipid composition, temperature, or pH affect the partition coefficient, it is important to better define the distribution of Tempo and Tempol for each experimental system. The equilibrium dialysis method (as described in Results) used in the present study is free of the above-mentioned limitations. Further elucidation of the mode of protection requires: (a) expanding the study to include other nitroxides having very high and very low partition coefficients as well as nitroxides of other basic molecular structure than the six-membered-ring molecules used here, and (b) studying the specific localization of the nitroxide along the phospholipid molecules, in order to assess the importance of this factor to the protective effect. The present results suggest that the protection of lipid vesicles against radiation-induced LPO is mediated by scavenging of reactive oxygen species formed in the aqueous phase, near the lipid bilayer. The nitroxides’ partial regeneration, i.e., catalytic antioxidant activity, and their lack of prooxidant activity make them superior to most conventional antioxidants.
/ 2b27 2426 Mp 1172 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage CONCLUSION
The results of the present study indicate that: (a) The higher the degree of unsaturation in the acyl chain, the greater is the degradation caused by irradiation. (b) The fully saturated fatty acids, palmitic acid (C16), and stearic acid (C18) showed no significant change in their levels. (c) Both Tempo and Tempol provided similar protection to acyl chain residues. (d) Nitroxides’ lipidbilayer/aqueous distribution is not validly represented by their n-octanol/saline partition coefficient. (e) The lipid-bilayer/aqueous partition coefficient of Tempo and Tempol can not be correlated with their protective effect. (f) The nitroxides appear to protect via a catalytic mode. Acknowledgements — This work was supported in part by the Szold Foundation, Jerusalem, Israel. We would like to acknowledge Dr. Z. Tochner and Dr. S. Rudman, Institute of Radiotherapy, Hadassah Hospital, Ein-Karem, Jerusalem, for their help in g-irradiation of our samples. The help of Mr. Sigmund Geller in editing the manuscript is acknowledged with pleasure.
15. 16.
17.
18. 19. 20.
21. 22. 23.
REFERENCES 1. Shmeeda, H. R.; Golden, E. B.; Barenholz, Y. Membrane lipids and aging. In: Shinitzky, M., ed. Biomembranes. Structural and functional aspects. Vol. 2: Weinheim: VCH; 1994:1–82. 2. Gregoriadis, G., ed. Liposomes as drug carriers. Chichester, UK: John Wiley & Sons Ltd; 1988. 3. Lasic, D. D. Liposomes: From physics to applications. Amsterdam: Elsevier Science Publishers B. V.; 1993. 4. Barenholz, Y.; Lasic, D. D., eds. Handbook of nonmedical applications of liposomes. Boca Raton, FL: CRC Press; 1996. 5. Holman, R. T., ed. Progress in the chemistry of fats and other lipids. New York: Pergamon Press; 1954. 6. Frankel, E. N. Lipid oxidation: mechanisms, products, and biological significance. J. Am. Oil Chem. Soc. 61:1908–1917; 1984. 7. Barber, A. A.; Bernheim, F. Lipid peroxidation: Its measurement, occurrence, and significance in animal tissues. Adv. Gerontol. Res. 2:355–403; 1967. 8. Lang, J. K.; Vigo-Pelfrey, C. Quality control of liposomal lipids with special emphasis on peroxidation of phospholipids and cholesterol. Chem. Phys. Lipids 64:19–29; 1993. 9. Fukuzawa, K.; Kishikawa, K.; Tadokoro, T.; Tokumura, A.; Tsukatani, H.; Gebicki, J. M. The effects of alpha-tocopherol on sitespecific lipid peroxidation induced by iron in charged micelles. Arch. Biochem. Biophys. 260:153–160; 1988. 10. Rachidi, S.; Coudray, C.; Baret, P.; Gelon, G.; Pierre, J. L.; Favier, A. Inhibition of lipid peroxidation by a new family of iron chelators. Comparison with desferrioxamine. Biol. Trace Elem. Res. 41:77–87; 1994. 11. Lichtenberg, D.; Barenholz, Y. Liposomes: Preparation, characterization, and preservation. In: Glick, D., ed. Methods of biochemical analysis. Vol. 33. New York: John Wiley & Sons, Inc.; 1988:337–462. 12. Burton, G. W.; Foster, D. O.; Perly, B.; Slater, T. F.; Smith, I. C. P.; Ingold, K. U. Biological antioxidants. Philos. Trans. R. Soc. Lond. [Biol.] 311:565–578; 1985. 13. Liebler, D. C.; Kling, D. S.; Reed, D. J. Antioxidant protection of phospholipid bilayers by alpha-tocopherol. Control of alphatocopherol status and lipid peroxidation by ascorbic acid and glutathione. J. Biol. Chem. 261:12114–12119; 1986. 14. Afzal, V.; Brasch, R. C.; Nitecki, D. E.; Wolff, S. Nitroxyl spin label contrast enhancers for magnetic resonance imaging. Studies
24. 25.
26.
27. 28. 29. 30. 31.
32. 33.
34.
35.
1173
of acute toxicity and mutagenesis. Invest. Radiol. 19:549–552; 1984. Ankel, E. G.; Lai, C. S.; Hopwood, L. E.; Zivkovic, Z. Cytotoxicity of commonly used nitroxide radical spin probes. Life Sci. 40:495–498; 1987. DeGraff, W. G.; Krishna, M. C.; Russo, A.; Mitchell, J. B. Antimutagenicity of a low molecular weight superoxide dismutase mimic against oxidative mutagens. Environ. Mol. Mutagen. 19:21–26; 1992. Reddan, J. R.; Sevilla, M. D.; Giblin, F. J.; Padgaonkar, V.; Dziedzic, D. C.; Leverenz, V.; Misra, I. C.; Peters, J. L. The superoxide dismutase mimic TEMPOL protects cultured rabbit lens epithelial cells from hydrogen peroxide insult. Exp. Eye. Res. 56:543–554; 1993. McConnell, H. M. In: Berliner, L. J., ed. Spin labeling: Theory and applications. New York: Academic Press; 1976:525. Kocherginsky, N.; Swartz, H. M. Nitroxide spin labels: Reactions in biology and chemistry. Boca Raton, FL: CRC Press; 1995. Swartz, H. M.; Sentjurc, M.; Morse, P. d. Cellular metabolism of water-soluble nitroxides: Effect on rate of reduction of cell/ nitroxide ratio, oxygen concentrations and permeability of nitroxides. Biochim. Biophys. Acta 888:82–90; 1986. Miura, Y.; Utsumi, H.; Hamada, A. Antioxidant activity of nitroxide radicals in lipid peroxidation of rat liver microsomes. Arch. Biochem. Biophys. 300:148–156; 1993. Nilsson, U. A.; Olsson, L. I.; Carlin, G.; Bylund, F. A. Inhibition of lipid peroxidation by spin labels. Relationships between structure and function. J. Biol. Chem. 264:11131–11135; 1989. Nilsson, U. A.; Carlin, G.; Bylund, F. A. The hydroxylamine OXANOH and its reaction product, the nitroxide OXANO, act as complementary inhibitors of lipid peroxidation. Chem. Biol. Interact. 74:325–342; 1990. Krishna, M. C.; Samuni, A. Nitroxides as antioxidants. Methods Enzymol. 234:580–589; 1994. Barenholz, Y.; Amselem, S. Quality control assays in the development and clinical use of liposome-based formulations. In: Gregoriadis, G., ed. Liposome technology. 2nd ed. Vol. I. Boca Raton, FL: CRC Press; 1993:527–616. Haran, G.; Cohen, R.; Bar, L. K.; Barenholz, Y. Transmembrane ammonium sulfate gradients in liposomes produce efficient and stable entrapment of amphipathic weak bases. Biochim. Biophys. Acta 1151:201–215; 1993. Bligh, E. G.; Dyer, W. J. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911–917; 1959. Rozantsev, E. G.; Sholle, V. D. Synthesis and reactions of stable nitroxyl radicals. I. Synthesis. Synthesis 190; 1971. Janero, D. R. Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and peroxidative tissue injury. Free Radic. Biol. Med. 9:515–540; 1990. Gutteridge, J. M. C.; Halliwell, B. The measurement and mechanism of lipid peroxidation in biological systems. Trends Biosci. 15:129–135; 1990. Barenholz, Y.; Cohen, T.; Korenstein, R.; Ottolenghi, M. Organization and dynamics of pyrene and pyrene lipids in intact lipid bilayers. Photo-induced charge transfer processes. Biophys. J. 60:110–124; 1991. Yoshida, Y.; Tsuchiya, J.; Niki, E. Interaction of alpha-tocopherol with copper and its effect on lipid peroxidation. Biochim. Biophys. Acta 1200:85–92; 1994. Goffman, T.; Cuscela, D.; Glass, J.; Hahn, S.; Krishna, C. M.; Lupton, G.; Mitchell, J. B. Topical application of nitroxide protects radiation-induced alopecia in guinea pigs. Int. J. Radiat. Oncol. Biol. Phys. 22:803–806; 1992. Hahn, S. M.; Wilson, L.; Krishna, C. M.; Liebmann, J.; DeGraff, W.; Gamson, J.; Samuni, A.; Venzon, D.; Mitchell, J. B. Identification of nitroxide radioprotectors. Radiat. Res. 132:87–93; 1992. Hahn, S. M.; Tochner, Z.; Krishna, C. M.; Glass, J.; Wilson, L.; Samuni, A.; Sprague, M.; Venzon, D.; Glatstein, E.; Mitchell, J. B. Tempol, a stable free radical, is a novel murine radiation protector. Cancer Res. 52:1750–1753; 1992.
/ 2b27 2426 Mp 1173 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1174
A. M. SAMUNI and Y. BARENHOLZ
36. Mitchell, J. B.; DeGraff, W.; Kaufman, D.; Krishna, M. C.; Samuni, A.; Finkelstein, E.; Ahn, M. S.; Hahn, S. M.; Gamson, J.; Russo, A. Inhibition of oxygen-dependent radiation-induced damage by the nitroxide superoxide dismutase mimic, tempol. Arch. Biochem. Biophys. 289:62–70; 1991. 37. Niki, E.; Takahashi, M.; Komuro, E. Antioxidant activity of vitamin E in liposomal membranes. Chem. Lett. (Jpn.) 9:1573–1576; 1986. 38. Wills, E. D. Effect of antioxidants on lipid peroxide formation in irradiated synthetic diets. Int. J. Radiat. Biol. 37:403–414; 1980. 39. Fukuzawa, K.; Takase, S.; Tsukatani, H. The effect of concentration on the antioxidant effectiveness of alpha-tocopherol in lipid peroxidation induced by superoxide free radicals. Arch. Biochem. Biophys. 240:117–120; 1985. 40. Stoyanovsky, D. A.; Kagan, V. E.; Packer, L. Iron binding to alpha-tocopherol-containing phospholipid liposomes. Biochem. Biophys. Res. Commun. 160:834–838; 1989. 41. Rekker, R. F.; Mannhold, R. Calculation of drug lipophilicity. New York: VCH; 1992. 42. McConnell, H. M.; Kenneth, L. W.; McFarland, B. G. The fraction of the lipid in a biological membrane that is in a fluid state: A spin label assay. Biochem. Biophys. Res. Commun. 47:273– 281; 1972. 43. Rosen, G. M.; Finkelstein, E.; Rauckman, E. J. A method for the detection of superoxide in biological systems. Arch. Biochem. Biophys. 215:367–378; 1982. 44. Fuchs, J.; Nitschmann, W. H.; Packer, L.; Hankovszky, O. H.; Hideg, K. pKa values and partition coefficients of nitroxide spin probes for membrane bioenergetics measurements. Free Radic. Res. Commun. 10:315–323; 1990.
45. Barclay, L. R. C.; Artz, J. D.; Mowat, J. J. Partitioning and antioxidant action of the water-soluble antioxidant, Trolox, between the aqueous and lipid phases of phosphatidylcholine membranes: 14C tracer and product studies. Biochim. Biophys. Acta 1237:77–85; 1995. 46. Wimley, W. C.; White, S. H. Membrane partitioning: Distinguishing bilayer effect from the hydrophobic effect. Biochemistry 32:6307–6312; 1993.
ABBREVIATIONS
EPC—egg phosphatidylcholine EPR—electron paramagnetic resonance GC—gas liquid chromatography LPO—lipid peroxidation PC—phosphatidyl choline PUFA—polyunsaturated fatty acids RT—room temperature SUV—small unilamellar vesicles Tempol — 4-hydroxy-2,2,6,6-tetramethylpiperidine-1oxyl Tempo—2,2,6,6-tetramethylpiperidine-1-oxyl Tempol-H — 4-hydroxy-2,2,6,6-tetramethyl-1hydroxypiperidine Tempo-H—2,2,6,6-tetramethyl-1-hydroxypiperidine
/ 2b27 2426 Mp 1174 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
PII S0891-5849(96)00509-6
Original Contribution STABLE NITROXIDE RADICALS PROTECT LIPID ACYL CHAINS FROM RADIATION DAMAGE Ayelet M. Samuni and Yechezkel Barenholz Department of Biochemistry, Hebrew University, Hadassah Medical School, Jerusalem 91120, Israel (Received 6 April 1996; Revised 12 August 1996; Accepted 23 September 1996)
Abstract—The present study focused on protective activity of two six-membered-ring nitroxide radicals, 2,2,6,6tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy-Tempo (Tempol), against radiation damage to acyl chain residues of egg phosphatidylcholine (EPC) of small unilamellar vesicles (SUV). SUV were g -irradiated (10–12 kGy) under air at ambient temperature in the absence and presence of nitroxides. Acyl chain composition of the phospholipids before and after irradiation was determined by gas chromatography. Both Tempo and Tempol effectively and similarly protected the acyl chains of EPC SUV, including the highly sensitive polyunsaturated acyl chains, C20:4, C22:5, and C22:6. The conclusions of the study are: (a) The higher the degree of unsaturation in the acyl chain, the greater is the degradation caused by irradiation. (b) The fully saturated fatty acids palmitic acid (C16) and stearic acid (C18) showed no significant change in their levels. (c) Both Tempo and Tempol provided similar protection to acyl chain residues. (d) Nitroxides’ lipid-bilayer/aqueous distribution is not validly represented by their n-octanol/ saline partition coefficient. (e) The lipid-bilayer/aqueous partition coefficient of Tempo and Tempol cannot be correlated with their protective effect. (f) The nitroxides appear to protect via a catalytic mode. Unlike common antioxidants, such as a-tocopherol, which are consumed under irradiation and are, therefore, less effective against high radiation dose, nitroxide radicals are restored and terminate radical chain reactions in a catalytic manner. Furthermore, nitroxides neither yield secondary radicals upon their reaction with radicals nor act as prooxidants. Not only are nitroxides self-replenished, but also their reduction products are effective antioxidants. Therefore, the use of nitroxides offers a powerful strategy to protect liposomes, membranes, and other lipid-based assemblies from radiation damage. q 1997 Elsevier Science Inc. Keywords—Spin labels, Lipid peroxidation, Liposomes, Antioxidants, SOD-mimic, Superoxide
carriers for a broad spectrum of agents in medical and nonmedical applications.2 – 4 Phosphatidylcholines (lecithins) and cholesterol are the major components of eukaryotic biological membranes, and of many lipid-based pharmaceutical formulations. Lecithins are a heterogeneous group of compounds whose constituent fatty acids vary in chain length and degree of unsaturation. Most naturally occurring lecithins (such as in egg yolk) contain substantial amounts of polyunsaturated fatty acids (PUFA), which are particularly susceptible to oxidation mediated by free radicals.5 – 8 The protection of phospholipid-based formulations is better controllable than that of membranes in vivo. Preventive and protective measures are generally taken to minimize oxidative damage of lipid-based emulsions
INTRODUCTION
Lipid oxidation, also called lipid peroxidation (LPO), involves various processes that lead to the formation of a broad spectrum of products and intermediates, including peroxides, alkoxides, aldehydes, and fatty acids. It has long been known as a problem in the preparation and preservation of lipid matrix and liposomal dispersions and formulations, and has been implicated in degeneration processes in aging and various pathological conditions.1 LPO is also a major obstacle to expanding the use of formulations that are based on lipid assemblies, such as emulsions and liposomes, as Address correspondence to: Ayelet M. Samuni, Department of Biochemistry, Hebrew University, Hadassah Medical School, Jerusalem 91120, Israel. 1165
/ 2b27 2426 Mp 1165 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1166
A. M. SAMUNI and Y. BARENHOLZ
and liposomal formulations. In addition to using the proper conditions of temperature, pH, and light, to minimize LPO, the traditional means of intervention include the use of chelating agents, such as Desferal, which neutralize redox-active metals, and lipophilic antioxidants, such as a-tocopherol (vitamin E), which act as a scavenger of organic free radicals.9,10 Many antioxidants are approved for use in food products and some have been used to preserve liposomes. No systematic studies to establish the ideal antioxidant have been carried out on liposome systems. The literature contains many disagreements as a result of differences in experimental details, which make it difficult to compare the various antioxidants. Unlike isotropic oil systems, mono- and bilayered assemblies like micelles, emulsions, liposomes, and biological membranes are anisotropic structures having a well-defined organization. Consequently, the distribution of antioxidant in the lipid assembly, the location of its active moiety in the anisotropic system and its lateral diffusion rate might be crucial to the antioxidative activity.11 Watersoluble antioxidants like ascorbic acid are reportedly less effective by themselves in preventing lipid peroxidation, although they improve protection by regenerating lipophilic antioxidants.12,13 A newer strategy that has been introduced employs stable nitroxide radicals to control degradation processes mediated by deleterious reactive species. Nitroxides are stable, nontoxic, nonimmunogenic,14 – 17 cyclic radicals, of diverse size, charge, lipophilicity, and permeability through biological membranes, which have been widely used as biophysical probes, spin labels, and contrast agents for nuclear magnetic resonance imaging.18 – 20 Inhibition of LPO in rat liver microsomes by both five- and six-membered-ring nitroxides has been previously reported.21 – 23 The antioxidant effects of both lipophilic and hydrophilic nitroxides, as well as of their respective hydroxylamines, indicate the operation of different mechanisms underlying the prevention of LPO.21 – 23 Other studies showed that nitroxides can protect cells, isolated organs, and whole animals against oxidative stress inflicted by diverse insults.24 Radical-mediated peroxidation processes, as in the case of g-irradiation, yield a variety of products, some of which appear only transiently, and undergo fast changes. This is the case for the fatty acid peroxidation indicators malonyl dialdehyde (MDA) and lipid-hydroperoxide (LOOH). As shown by Lang and Vigo-Pelfrey, the accumulation of MDA and LOOH does not validly reflect fatty acid disappearance. In contrast, a quantitative determination of individual fatty acids (using GC) or cholesterol (using HPLC) enables an accurate and valid evaluation of membrane lipid degra-
dation.8 The present study focused on the damage inflicted to acyl chain residues of egg phosphatidylcholine (EPC) by g-irradiation, and the protection provided by nitroxide radicals. g-Irradiation, selected as the means of inducing damage, is a well defined procedure, allowing good control over the dose administered, onset and termination of reaction, uniform exposure to the insult, the radical species formed, and does not require the addition of components to the test sample. The protection by 2,2,6,6-tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (Tempol)
of acyl chains of small unilamellar vesicles (SUV) prepared from EPC, from radiation, and its relevance to the nitroxide concentration in the lipid bilayer and in the vesicle aqueous phase, were studied.
MATERIALS AND METHODS
Materials Egg phosphatidylcholine (containing 95.6% PC, 3.1% sphingomyelin, 0.5% lyso PC and no other lipids, containing 0.12% DL-a-tocopherol) was purchased from Lipoid KG (Ludwigshafen, Germany); L-a-phosphatidylcholine (egg) (ú99%) from Avanti Polar Lipids (Alabaster, AL, USA); n-octyl alcohol from Riedelde haen AG, Seelze-Hannover, Germany; sterile nonpyrogenic saline from Travenol, Israel; 2,2,6,6-tetramethylpiperidine-1-oxyl (Tempo) and 4-hydroxy2,2,6,6-tetramethylpiperidine-1-oxyl (Tempol) from Aldrich, Milwaukee, WI; Meth-Prep II kit from Alltech Associates, Deerfield, IL. All other chemicals were of analytical grade. Bidistilled water used was also deionized by millipore column.
Determination of phospholipid concentration Phospholipid concentration of the liposome preparation was determined using a modification of the Bartlett procedure.25
/ 2b27 2426 Mp 1166 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
Particle size determination The distribution of the particle size in the liposomal dispersion was measured by photon correlation spectroscopy using a Coulter Model N4 SD apparatus.25
Preparation of liposomes SUV liposomes were prepared from EPC, 12% in sterile nonpyrogenic water, by high-pressure homogenization as described elsewhere,26 using a two-stage high-pressure homogenizer Gaulin LAB 60 (APV Gaulin, Hilversum, Holland). Vesicle size distribution was unimodal, with mean vesicle size of 25 { 5 nm.
Ionizing radiation Standard steady-state radiochemical technique was employed, using a 60Co g-source at the radiotherapy unit, Hadassah Hospital, Ein-Karem, Jerusalem, with a dose rate of 10 Gyimin01. Samples were prepared in test tubes and irradiated at ambient temperature, without stirring, under air, protected from light. Radiation dose varied between 10 and 12 kGy due to experimental restraints. The exact dose is stated for each specific case.
Determination of acyl chain peroxidation Acyl chain composition of the phospholipids before and after irradiation was determined as described by Barenholz and Amselem25 and summarized briefly as follows: after a Bligh and Dyer extraction,27 the lower (chloroform-rich) phase was transferred to a small glass bottle, and evaporated under N2 to complete dryness. To transmethylate the acyl chains of the phospholipids, 50 ml of toluene and 20 ml of Meth-Prep II were added to the extracted lipid. The samples were left for 30 min at RT, and injected (volume of 2 ml) into a Perkin– Elmer AutoSystem gas chromatography (GC) and Autosampler, using a 6-ft 10% Silar 10C column (Alltech), dry N2 as the carrier gas, and a flame ionization detection. The initial temperature of the run was 1407C for 5 min, then the oven temperature was raised at a rate of 57C/min up to 2407C, and then kept there for 5 min. Methyl esters were identified by comparing the retention times with those of known standards. Palmitic acid (C16), a saturated fatty acid, which has been previously found resistant to peroxidation by g-irradiation, was selected as a reference for the determination of the extent of degradation in other acyl chains of EPC.
1167
Electron paramagnetic resonance (EPR) measurements EPR spectrometry was employed to detect and follow the nitroxide free radicals, using a JES-RE3X ESR spectrometer (JEOL Co., Japan). Samples were drawn by a micropippette into a gas-permeable Teflon capillary of 0.81 mm inner diameter, 0.05 mm wall thickness, and 15 cm length (Zeus Industrial Products, Raritan, NJ). Each capillary was folded twice, inserted into a 2.5 mm i.d. quartz tube open at both ends, and placed in the EPR cavity. EPR spectra were recorded with center field set at 3361 G, 100 kHz modulation frequency, 1 G modulation amplitude, and nonsaturating microwave power. Nitroxides decay in biological systems predominantly through a one-electron reduction yielding the respective cyclic hydroxylamines.20 For determination of the total concentration of nitroxide / hydroxylamine, the hydroxylamine was oxidized by 1 mM ferricyanide or 0.1 M H2O2 at pH 12. 28 Nitroxide lipophilicity, n-octanol-to-saline partition To assess nitroxide lipophilicity, the partition between n-octanol and saline was studied and the n-octanol/saline partition coefficient, Koct/aq, of the nitroxide was determined. n-Octanol and saline (1/1; v/v) were thoroughly mixed and a known amount of nitroxide (Tempo or Tempol) was added. Samples were prepared at various initial concentrations of nitroxide. The mixture was vortexed for 5 min (shaking the samples for 24 h gave the same results) and then left for 1 h until separation of phases. The samples were centrifuged for 10 min (7000 1 g) to facilitate separation of the two phases. An aliquot of each of the two phases was sampled and the intensities of the respective EPR signals were compared. For calibration, both Tempo and Tempol were dissolved in n-octanol and in saline at various concentrations in the range of 0.1–1.0 mM, their EPR signals were determined, and the respective calibration curves were obtained. Partition of nitroxide between lipid bilayer and saline Partitioning of the nitroxides between the liposome lipid bilayer and the aqueous phase was measured using a two-compartment Lucite cuvette specially designed for equilibrium dialysis. Each compartment has a volume of 0.3 ml, and the two compartments were separated by a dialysis membrane of molecular weight cut of 12–14,000 daltons. One compartment contained the liposomal dispersion and the other compartment contained saline. The samples were incubated under con-
/ 2b27 2426 Mp 1167 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1168
A. M. SAMUNI and Y. BARENHOLZ
tinuous shaking, at RT, to allow the dialysismembrane-permeable nitroxide to equilibrate between the two compartments. Following 24 h incubation, samples from the liposome-free and liposome-containing compartments were taken, scanned in the EPR spectrometer, and the respective intensities Cliposome and Caq of the EPR signal of nitroxide in each compartment were compared. Because nitroxide EPR signals in the lipid and aqueous phases are different, and the differences are dependent on the specific nitroxide, the intensities of the EPR signal of Tempo and Tempol at various lipid concentrations were measured, and calibration curves were constructed, by means of which the actual nitroxide concentrations in each compartment were calculated. Sample preparation Samples, prepared in saline pH 4.5, or in 10 mM phosphate buffer pH 7.4, contained 5 mM or 20 mM of SUV phospholipid, with an average particle size of 25 { 5 nm. Various radiation doses were tested, and significant loss of PUFA was induced using á 10 kGy; therefore, the samples were g-irradiated with a dose of 10–12 kGy, under air, at ambient temperature, in the absence and presence of nitroxide. Using EPC without vitamin E gave similar results. In all experiments, the preparations and measurements of samples were carried out under air, at RT, with minimal exposure to light.
mination of degradation in other acyl chains of EPC (Fig. 1). The results show increasing damage with decreasing saturation of the fatty acid, reaching over 80% loss for the most polyunsaturated fatty acid, C22:6 (docosahexanoic). The extent of damage increased with the decrease in phospholipid concentration. Addition of formate (0.1 M), which under air converts iOH, iH, and e0aq into secondary Oi20 radicals, in the absence or the presence of 5–50 mM Cu(II) or Fe(II) did not potentiate the damage. Formate alone even slightly decreased the damage. Protection by nitroxides. In nonirradiated liposomal dispersions no effect of 5 mM nitroxide on liposome size or chemical composition was found. Both nitroxides protected the unsaturated acyl chains in a concentration-dependent manner against radiation damage induced in phosphatidylcholine (Fig. 2). Even the most sensitive polyunsaturated fatty acids such as C20:4, C22:5, and C22:6 were fully protected by 5mM nitroxide, whereas 1 mM nitroxide provided partial protection, preventing about 50% of the damage (Fig. 2).
RESULTS
Acyl chain peroxidation PUFA degradation. In studies of LPO, the accumulation of MDA is generally used to assess the extent of damage. The validity of this widely used procedure has been questioned8,29,30 considering various artefacts associated with this assay:25,29,30 (a) most of the MDA detected is generated during the test itself; (b) added metal ions, H2O2, antioxidants, and chelating agents influence peroxide decomposition during the assay itself; and (c) the type and concentration of acid added also influence the amount of thiobarbituric acid reactive substance formed (TBAR). In the present study, the degradation of the acyl chain residues of the phospholipids following 10 kGy irradiation was measured using GC, which enables an accurate quantification of each of the acyl chains without interference of other oxidation products (such as those of cholesterol). Palmitic acid (C16), a saturated fatty acid, which is resistant to peroxidation by g-irradiation, served as an internal reference for the deter-
Fig. 1. Effect of g-irradiation on composition of fatty acids in EPC SUV. Liposomal dispersion of EPC SUV in saline was g-irradiated with 10 kGy at RT. Fatty acid composition was determined after methyl esterification followed by GC separation. The new levels of fatty acids, related to palmitic acid, which served as internal standard, are presented as a fraction of their original levels. Upper— 5 mM phospholipid; Lower—20 mM phospholipid.
/ 2b27 2426 Mp 1168 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
1169
coefficient values, Koct/aq determined for Tempo and Tempol are listed in Table 1. Nitroxide lipid-bilayer/saline partition
Fig. 2. Tempo inhibits radiation damage induced in PUFA. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of several concentrations of 0.1– 5.0 mM Tempo. Fatty acid composition was determined after methyl esterification followed by GC separation. The residual levels of six unsaturated fatty acids, related to palmitic acid, which served as internal standard, are presented as a fraction of their original levels. Inset: Comparison of the inhibitory effect provided by Tempo and Tempol against radiation damage induced in PUFA. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of 1 mM Tempo or Tempol.
Both nitroxides had a similar protective effect, with no statistically significant difference (using the Mann– Whitney U-test), as will be discussed below. Nitroxides’ partition between octanol/saline and liposomal-bilayer/saline Lipophilicities of the nitroxides were evaluated by determining the coefficient of partition between the lipophilic and aqueous compartments.
Coefficients of nitroxide partition between liposomal lipid bilayer assembly and aqueous phase were determined using a two-compartment cuvette specially designed for equilibrium dialysis. This method excludes the need to separate the lipid bilayer from the aqueous phase in a liposomal dispersion. Following 24 h incubation under continuous shaking at RT, samples from the liposome-free and liposome-containing compartments were taken and scanned in the EPR spectrometer. The nitroxide concentration Cliposome and Caq in the two compartments were calculated from their EPR signals according to their respective calibration curves (in the different media). The experiments were repeated using various lipid concentrations, the values of the ratio Cliposome/Caq were plotted vs. the volume fraction, f, occupied by the lipid bilayer (Fig. 3). The volume fraction f was calculated using partial specific volume (vV ) of 0.984 mlig01 for EPC SUVs.31 The experimental values of Cliposome/Caq increased linearly with f, obeying the expression: Cliposome /C aq Å (1 0 f) / fiClipid /C aq Å (1 0 f) / fiK lipid/aq
from which the partition coefficients, Klipid/aq, were calculated. Lipid-bilayer/saline partition coefficients determined for Tempo and Tempol are listed in Table 1. As can be calculated using the respective Klipid/aq values, the lipid phase in a liposome dispersion of 5 mM EPC contains 10.45% and 1.3% of Tempo and Tempol, respectively. Figure 4 demonstrates the degree of protection provided the acyl chain 20:4 as a function of the nitroxides’ (Tempo and Tempol) concentrations in the aqueous and lipid phases. Nitroxides’ fate upon g-irradiation
Nitroxide octanol/saline partition The partition between octanol and water was determined using EPR spectrometry of the nitroxide concentration in the two phases. Octanol/saline partition
EPR spectra of irradiated and nonirradiated samples containing 5 mM nitroxide were measured and compared in order to determine the residual fraction of nitroxide remaining following irradiation in the absence
Table 1. Partition Coefficients of Tempo and Tempol Between Octanol and Saline, Koct/aq, and Between Lipid-Bilayer and Saline, Klipid/aq Nitroxide
Octanol/Saline Koct/aq
Lipid Bilayer/Saline Klipid/aq
Koct/aq:Klipid/aq
Tempo Tempol
85.15 { 9.49 3.70 { 0.46
31.09 { 0.71 3.50 { 0.41
2.74 1.06
/ 2b27 2426 Mp 1169 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1170
A. M. SAMUNI and Y. BARENHOLZ
Fig. 3. Tempo and Tempol lipid-bilayer to saline partition coefficients. Partition coefficients of Tempo and Tempol between lipid-bilayer and saline were measured using equilibrium dialysis double compartment cell. The concentration of nitroxide in lipid-free (Caq) and lipid-containing (Cliposome) solution in two chambers, separated by a semipermeable membrane, were measured by EPR and compared. The lipid concentration was varied and the ratio Cliposome/Caq was determined. The partition coefficient, Klipid/aq, was calculated considering the partial specific volume of EPC SUV, at various lipid concentrations, using the expression: Cliposome/Caq Å (1 0 f) / fKlipid/aq.
and the presence of 20 mM SUV. The results showed that the residual nitroxide (Tempo or Tempol) was dependent on the irradiation dose and on the absence or presence of the lipid. The results are listed in Table 2. To determine the total level of nitroxide plus its respective hydroxylamine, the sample was incubated for 1 h with 0.1 M H2O2 at pH 12 or with 1 mM ferricyanide. No increase in the EPR signal of Tempol or Tempo was found, indicating that no appreciable Tempol-H or Tempo-H, the respective hydroxylamines, were accumulated upon irradiation. DISCUSSION
A common strategy of protection against LPO employs reducing agents that act as preventive and chain-
Fig. 4. Protection of acyl chain C20:4 (arachidonic acid) from radiation as a function of the nitroxide concentrations in the aqueous and lipid phases. Liposomal dispersion of 20 mM EPC SUV in saline was g-irradiated at RT with 10 kGy, in the presence of several concentrations of nitroxide. The extent (%) of protection provided by Tempo (squares) and Tempol (circles) to the acyl chain is plotted as a function of their concentration in the two phases: aqueous phase (open symbols) and lipid bilayer (solid symbols).
breaking antioxidants. However, their efficacy is limited because they (a) are being depleted; (b) give rise to secondary radicals which may be deleterious themselves; and (c) may act as prooxidants.32 Nitroxides, on the other hand, act catalytically and are self-replenished, without any prooxidative effect. They can protect cells and the whole animal from radiation33–36 and represent a new class of nonthiol aerobic radioprotectors. The reactions of nitroxides, being radicals themselves, as antioxidants and radioprotectors do not yield secondary radicals, but rather terminate the radical chain reactions. Nitroxides have been shown to react with peroxyl radicals and inhibit lipid peroxidation,21,22 and their activity has been attributed to their respective hydroxylamines.
Nitroxides inhibit acyl chain degradation Hydrophilic antioxidants were previously reported to be less effective in protection against LPO induced by g-irradiation than lipophilic ones. Their effect was attributed to a regeneration of lipophilic antioxi-
Table 2. Residual Fraction of Nitroxide Remaining Following g-Irradiation in the Absence or Presence of 20 mM SUV 12 kGy
32 kGy
Nitroxide
no SUV
/SUV
no SUV
/SUV
Tempo (5 mM) Tempol (5 mM)
26 { 3 26 { 4
34 { 6 38 { 4
0.4 { 0.1 6.8 { 0.1
17 { 4 17 { 4
/ 2b27 2426 Mp 1170 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage
dants.12,13 The rationale behind the use of lipophilic antioxidants is their selective incorporation and retention in the membrane (or lipid-bilayer), thus being in close proximity to the site of damage. The present study evaluates the capacity of the amphipathic nitroxides to protect against g-irradiation-induced LPO (PUFA degradation). We tried to address major issues in optimizing the selection of protecting agents against g-irradiation, especially to assess to what extent their lipophilicity and concentration in the aqueous phase and in the membrane affect their protective effect. We demonstrated that the protective effect is dependent on the concentration of the nitroxide(s) in the aqueous phase, achieving for both Tempo and Tempol complete protection at 5 mM (for 20 mM EPC). As will be discussed below, the protective effect is neither related to the lipophilicity of the nitroxide (Tempo @ Tempol) as assessed by Koct/aq, nor by their concentration in the EPC lipid bilayer (Tempo ú Tempol). The results presented here suggest that for antioxidants that lack prooxidant effect, and that can be regenerated during the oxidation, low concentration in the membrane is not an obstacle to the protection against g-irradiation oxidative damage, as long as sufficient concentration in the aqueous phase can be achieved. The protective effect is concentration dependent, as can be seen in Figs. 2 and 4, achieving complete inhibition of PUFA degradation at 5 mM of Tempo (or Tempol). These results suggest the usefulness of watersoluble antioxidants compared to lipophilic antioxidants, and emphasize the need to further investigate the modes of damage and protection. Deleterious species, protective agents, and mechanism Our results showing no significant effect of formate on the irradiation damage suggest that iOH and Oi20 similarly contribute toward damage of lipid acyl chains during g-irradiation. On the other hand, iOH and Oi20 contribute differently toward nitroxide destruction. While the removal of Oi20 radicals by nitroxide is a catalytic process that does not affect its concentration, the abstraction of H by hydroxyls causes an irreversible destruction of the nitroxide. This explains the lower loss of nitroxide when irradiated in the presence of EPC SUV. The inhibition of nitroxide depletion in the presence of EPC SUV, indicates that its destruction is due to nonspecific reaction damage by the radiolytically formed hydroxyls, rather than its reaction with phospholipid-derived reactive species such as Li, LOi, or LOOi. The antioxidant activity of nitroxides has been previously demonstrated using rat liver microsomes and related largely to their respective cyclic hydroxyl-
1171
amines.21,22 In the presence of active reducing enzymatic systems the nitroxides are rapidly reduced to their respective hydroxylamines, Tempo-H and Tempol-H. In such cases, the protective species might be the hydroxylamine. In the present work the situation is different because the predominant species was found to be solely the nitroxide itself. Also, upon addition of Tempol-H to the test samples, no hydroxylamine was left following irradiation. Instead, Tempol was formed, indicating that Tempol-H was oxidized to Tempol. Both the nitroxide and its hydroxylamine can reduce carbon-centered and oxygen-centered lipid radicals and break the radical chain reaction:
Reactions 1 and 2 yield the nitroxide and its oxidized form, the oxo-ammonium cation, respectively. The latter can be rapidly reduced to its respective nitroxide through a diffusion-controlled reaction with superoxide, having k3 Å 1.5i1010 M01s01, as previously reported.19
Evidently, Oi20 radicals are removed also through reaction with the nitroxide itself:
/ 2b27 2426 Mp 1171 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1172
A. M. SAMUNI and Y. BARENHOLZ
This mechanism accounts for the replenishment of nitroxide with a concurrent removal of Oi20. While protection by nitroxides of lipids against iOH-induced damage is achieved in a suicidal mode that results with their progressive depletion, the mechanism underlying their protective activity against Oi20 induced damage and the breaking of radical chain reaction is catalytic. It is important to compare the nitroxides to traditional antioxidants. Among the lipophilic antioxidants most commonly used to inhibit LPO are tocopherols, and in particular a-tocopherol, a reactions chain breaking agent. Although the mechanism underlying the protective effect of tocopherols in membranes has not been fully elucidated, it is clear that they act in a stoichiometric fashion and are consumed while scavenging and preventing lipid peroxidation.37,38 Obviously, their concentration decreases upon exposure to a stress; therefore, they cannot protect against high doses of g-irradiation. Vitamin E was also shown to reduce transition metal ions, thus enhancing the Fenton reaction and causing a prooxidant rather than antioxidant effect.32 This constitutes another potential drawback in the use of vitamin E as antioxidant. In previous studies, 0.2–1.0 mol% of a-tocopherol (based on phospholipid content) effectively inhibited LPO.13,39,40 Wills reported38 an effective inhibition by 0.005–0.01% vitamin E of LPO in herring oil and starch mixtures irradiated with doses as high as 10 kGy. However, in the present study, 0.12 mol% DL-a-tocopherol in the EPC SUV did not prevent radiationinduced oxidation of EPC PUFA (Fig. 1). Effect of nitroxide lipophilicity and concentration in the membranes The lipophilicity may have a dual effect on the protective activity of an antioxidant against LPO: by (a) dictating its concentration at the site of LPO in the lipid bilayer, as expressed more validly by the bilayer/ aqueous partition coefficient (Klipid/aq) or (b) direct involvement in the antioxidant mechanism of action. The common way to describe lipophilicity is by the oil/aqueous (heptane/aqueous) and n-octanol/aqueous partition coefficient.41 For this reason, many studies on antioxidant activity of nitroxides consider the coefficients determined for partition between water and n-octanol.19,22,42 – 44 Our results show that Tempo’s partitions between n-octanol/saline and lipid-bilayer/saline greatly differ (Çthreefold, see Table 1). Not only does Koct/aq of a given nitroxide differ from its Klipid/aq, but also the ratio Koct/aq/Klipid/aq differs between Tempo and Tempol (Table 1). Koct/aq and Klipid/aq for Tempo and Tempol indicate that the Koct/aq value inaccurately de-
fines the nitroxide concentration in the bilayer. These results are in good agreement with recent studies on other molecules.45,46 This discrepancy is related to the large differences between bulk isotropic systems, such as oil or n-octanol, and the anisotropic microscopic environment of the lipid bilayer.46 This difference remains even when the dielectric constants of the bulk and the bilayer region, in which the antioxidant is localized, are matched (i.e., n-octanol and the bilayer interface region). According to their respective lipid bilayer/saline partition coefficients (3.5 and 31), Tempol and Tempo would have been anticipated to exert different protective effects. In fact, they provided similar protection against radiation-induced LPO. This finding indicates that both the large difference (23-fold) in Tempo and Tempol values of Koct/aq (lipophilicity) and lipid/aq partition coefficient (bilayer concentration), are not correlated with the protective effects of Tempo and Tempol, while their concentration in the aqueous phase is correlated with the antioxidant efficacy (Fig. 4). In previous attempts to determine the partition of nitroxide between aqueous and liposomal compartments, multilamellar vesicle (MLV) dispersions containing nitroxide were centrifuged and the EPR spectra of both the pellet and supernatant were compared.21 However, such an approach is less accurate due to the large interstitial volume of the MLV pellet. Also, small liposomes present in the supernatant may overestimate the level of nitroxide in the aqueous phase. Because variables such as lipid composition, temperature, or pH affect the partition coefficient, it is important to better define the distribution of Tempo and Tempol for each experimental system. The equilibrium dialysis method (as described in Results) used in the present study is free of the above-mentioned limitations. Further elucidation of the mode of protection requires: (a) expanding the study to include other nitroxides having very high and very low partition coefficients as well as nitroxides of other basic molecular structure than the six-membered-ring molecules used here, and (b) studying the specific localization of the nitroxide along the phospholipid molecules, in order to assess the importance of this factor to the protective effect. The present results suggest that the protection of lipid vesicles against radiation-induced LPO is mediated by scavenging of reactive oxygen species formed in the aqueous phase, near the lipid bilayer. The nitroxides’ partial regeneration, i.e., catalytic antioxidant activity, and their lack of prooxidant activity make them superior to most conventional antioxidants.
/ 2b27 2426 Mp 1172 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
Nitroxide radicals protect lipid acyl chains from damage CONCLUSION
The results of the present study indicate that: (a) The higher the degree of unsaturation in the acyl chain, the greater is the degradation caused by irradiation. (b) The fully saturated fatty acids, palmitic acid (C16), and stearic acid (C18) showed no significant change in their levels. (c) Both Tempo and Tempol provided similar protection to acyl chain residues. (d) Nitroxides’ lipidbilayer/aqueous distribution is not validly represented by their n-octanol/saline partition coefficient. (e) The lipid-bilayer/aqueous partition coefficient of Tempo and Tempol can not be correlated with their protective effect. (f) The nitroxides appear to protect via a catalytic mode. Acknowledgements — This work was supported in part by the Szold Foundation, Jerusalem, Israel. We would like to acknowledge Dr. Z. Tochner and Dr. S. Rudman, Institute of Radiotherapy, Hadassah Hospital, Ein-Karem, Jerusalem, for their help in g-irradiation of our samples. The help of Mr. Sigmund Geller in editing the manuscript is acknowledged with pleasure.
15. 16.
17.
18. 19. 20.
21. 22. 23.
REFERENCES 1. Shmeeda, H. R.; Golden, E. B.; Barenholz, Y. Membrane lipids and aging. In: Shinitzky, M., ed. Biomembranes. Structural and functional aspects. Vol. 2: Weinheim: VCH; 1994:1–82. 2. Gregoriadis, G., ed. Liposomes as drug carriers. Chichester, UK: John Wiley & Sons Ltd; 1988. 3. Lasic, D. D. Liposomes: From physics to applications. Amsterdam: Elsevier Science Publishers B. V.; 1993. 4. Barenholz, Y.; Lasic, D. D., eds. Handbook of nonmedical applications of liposomes. Boca Raton, FL: CRC Press; 1996. 5. Holman, R. T., ed. Progress in the chemistry of fats and other lipids. New York: Pergamon Press; 1954. 6. Frankel, E. N. Lipid oxidation: mechanisms, products, and biological significance. J. Am. Oil Chem. Soc. 61:1908–1917; 1984. 7. Barber, A. A.; Bernheim, F. Lipid peroxidation: Its measurement, occurrence, and significance in animal tissues. Adv. Gerontol. Res. 2:355–403; 1967. 8. Lang, J. K.; Vigo-Pelfrey, C. Quality control of liposomal lipids with special emphasis on peroxidation of phospholipids and cholesterol. Chem. Phys. Lipids 64:19–29; 1993. 9. Fukuzawa, K.; Kishikawa, K.; Tadokoro, T.; Tokumura, A.; Tsukatani, H.; Gebicki, J. M. The effects of alpha-tocopherol on sitespecific lipid peroxidation induced by iron in charged micelles. Arch. Biochem. Biophys. 260:153–160; 1988. 10. Rachidi, S.; Coudray, C.; Baret, P.; Gelon, G.; Pierre, J. L.; Favier, A. Inhibition of lipid peroxidation by a new family of iron chelators. Comparison with desferrioxamine. Biol. Trace Elem. Res. 41:77–87; 1994. 11. Lichtenberg, D.; Barenholz, Y. Liposomes: Preparation, characterization, and preservation. In: Glick, D., ed. Methods of biochemical analysis. Vol. 33. New York: John Wiley & Sons, Inc.; 1988:337–462. 12. Burton, G. W.; Foster, D. O.; Perly, B.; Slater, T. F.; Smith, I. C. P.; Ingold, K. U. Biological antioxidants. Philos. Trans. R. Soc. Lond. [Biol.] 311:565–578; 1985. 13. Liebler, D. C.; Kling, D. S.; Reed, D. J. Antioxidant protection of phospholipid bilayers by alpha-tocopherol. Control of alphatocopherol status and lipid peroxidation by ascorbic acid and glutathione. J. Biol. Chem. 261:12114–12119; 1986. 14. Afzal, V.; Brasch, R. C.; Nitecki, D. E.; Wolff, S. Nitroxyl spin label contrast enhancers for magnetic resonance imaging. Studies
24. 25.
26.
27. 28. 29. 30. 31.
32. 33.
34.
35.
1173
of acute toxicity and mutagenesis. Invest. Radiol. 19:549–552; 1984. Ankel, E. G.; Lai, C. S.; Hopwood, L. E.; Zivkovic, Z. Cytotoxicity of commonly used nitroxide radical spin probes. Life Sci. 40:495–498; 1987. DeGraff, W. G.; Krishna, M. C.; Russo, A.; Mitchell, J. B. Antimutagenicity of a low molecular weight superoxide dismutase mimic against oxidative mutagens. Environ. Mol. Mutagen. 19:21–26; 1992. Reddan, J. R.; Sevilla, M. D.; Giblin, F. J.; Padgaonkar, V.; Dziedzic, D. C.; Leverenz, V.; Misra, I. C.; Peters, J. L. The superoxide dismutase mimic TEMPOL protects cultured rabbit lens epithelial cells from hydrogen peroxide insult. Exp. Eye. Res. 56:543–554; 1993. McConnell, H. M. In: Berliner, L. J., ed. Spin labeling: Theory and applications. New York: Academic Press; 1976:525. Kocherginsky, N.; Swartz, H. M. Nitroxide spin labels: Reactions in biology and chemistry. Boca Raton, FL: CRC Press; 1995. Swartz, H. M.; Sentjurc, M.; Morse, P. d. Cellular metabolism of water-soluble nitroxides: Effect on rate of reduction of cell/ nitroxide ratio, oxygen concentrations and permeability of nitroxides. Biochim. Biophys. Acta 888:82–90; 1986. Miura, Y.; Utsumi, H.; Hamada, A. Antioxidant activity of nitroxide radicals in lipid peroxidation of rat liver microsomes. Arch. Biochem. Biophys. 300:148–156; 1993. Nilsson, U. A.; Olsson, L. I.; Carlin, G.; Bylund, F. A. Inhibition of lipid peroxidation by spin labels. Relationships between structure and function. J. Biol. Chem. 264:11131–11135; 1989. Nilsson, U. A.; Carlin, G.; Bylund, F. A. The hydroxylamine OXANOH and its reaction product, the nitroxide OXANO, act as complementary inhibitors of lipid peroxidation. Chem. Biol. Interact. 74:325–342; 1990. Krishna, M. C.; Samuni, A. Nitroxides as antioxidants. Methods Enzymol. 234:580–589; 1994. Barenholz, Y.; Amselem, S. Quality control assays in the development and clinical use of liposome-based formulations. In: Gregoriadis, G., ed. Liposome technology. 2nd ed. Vol. I. Boca Raton, FL: CRC Press; 1993:527–616. Haran, G.; Cohen, R.; Bar, L. K.; Barenholz, Y. Transmembrane ammonium sulfate gradients in liposomes produce efficient and stable entrapment of amphipathic weak bases. Biochim. Biophys. Acta 1151:201–215; 1993. Bligh, E. G.; Dyer, W. J. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911–917; 1959. Rozantsev, E. G.; Sholle, V. D. Synthesis and reactions of stable nitroxyl radicals. I. Synthesis. Synthesis 190; 1971. Janero, D. R. Malondialdehyde and thiobarbituric acid-reactivity as diagnostic indices of lipid peroxidation and peroxidative tissue injury. Free Radic. Biol. Med. 9:515–540; 1990. Gutteridge, J. M. C.; Halliwell, B. The measurement and mechanism of lipid peroxidation in biological systems. Trends Biosci. 15:129–135; 1990. Barenholz, Y.; Cohen, T.; Korenstein, R.; Ottolenghi, M. Organization and dynamics of pyrene and pyrene lipids in intact lipid bilayers. Photo-induced charge transfer processes. Biophys. J. 60:110–124; 1991. Yoshida, Y.; Tsuchiya, J.; Niki, E. Interaction of alpha-tocopherol with copper and its effect on lipid peroxidation. Biochim. Biophys. Acta 1200:85–92; 1994. Goffman, T.; Cuscela, D.; Glass, J.; Hahn, S.; Krishna, C. M.; Lupton, G.; Mitchell, J. B. Topical application of nitroxide protects radiation-induced alopecia in guinea pigs. Int. J. Radiat. Oncol. Biol. Phys. 22:803–806; 1992. Hahn, S. M.; Wilson, L.; Krishna, C. M.; Liebmann, J.; DeGraff, W.; Gamson, J.; Samuni, A.; Venzon, D.; Mitchell, J. B. Identification of nitroxide radioprotectors. Radiat. Res. 132:87–93; 1992. Hahn, S. M.; Tochner, Z.; Krishna, C. M.; Glass, J.; Wilson, L.; Samuni, A.; Sprague, M.; Venzon, D.; Glatstein, E.; Mitchell, J. B. Tempol, a stable free radical, is a novel murine radiation protector. Cancer Res. 52:1750–1753; 1992.
/ 2b27 2426 Mp 1173 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426
1174
A. M. SAMUNI and Y. BARENHOLZ
36. Mitchell, J. B.; DeGraff, W.; Kaufman, D.; Krishna, M. C.; Samuni, A.; Finkelstein, E.; Ahn, M. S.; Hahn, S. M.; Gamson, J.; Russo, A. Inhibition of oxygen-dependent radiation-induced damage by the nitroxide superoxide dismutase mimic, tempol. Arch. Biochem. Biophys. 289:62–70; 1991. 37. Niki, E.; Takahashi, M.; Komuro, E. Antioxidant activity of vitamin E in liposomal membranes. Chem. Lett. (Jpn.) 9:1573–1576; 1986. 38. Wills, E. D. Effect of antioxidants on lipid peroxide formation in irradiated synthetic diets. Int. J. Radiat. Biol. 37:403–414; 1980. 39. Fukuzawa, K.; Takase, S.; Tsukatani, H. The effect of concentration on the antioxidant effectiveness of alpha-tocopherol in lipid peroxidation induced by superoxide free radicals. Arch. Biochem. Biophys. 240:117–120; 1985. 40. Stoyanovsky, D. A.; Kagan, V. E.; Packer, L. Iron binding to alpha-tocopherol-containing phospholipid liposomes. Biochem. Biophys. Res. Commun. 160:834–838; 1989. 41. Rekker, R. F.; Mannhold, R. Calculation of drug lipophilicity. New York: VCH; 1992. 42. McConnell, H. M.; Kenneth, L. W.; McFarland, B. G. The fraction of the lipid in a biological membrane that is in a fluid state: A spin label assay. Biochem. Biophys. Res. Commun. 47:273– 281; 1972. 43. Rosen, G. M.; Finkelstein, E.; Rauckman, E. J. A method for the detection of superoxide in biological systems. Arch. Biochem. Biophys. 215:367–378; 1982. 44. Fuchs, J.; Nitschmann, W. H.; Packer, L.; Hankovszky, O. H.; Hideg, K. pKa values and partition coefficients of nitroxide spin probes for membrane bioenergetics measurements. Free Radic. Res. Commun. 10:315–323; 1990.
45. Barclay, L. R. C.; Artz, J. D.; Mowat, J. J. Partitioning and antioxidant action of the water-soluble antioxidant, Trolox, between the aqueous and lipid phases of phosphatidylcholine membranes: 14C tracer and product studies. Biochim. Biophys. Acta 1237:77–85; 1995. 46. Wimley, W. C.; White, S. H. Membrane partitioning: Distinguishing bilayer effect from the hydrophobic effect. Biochemistry 32:6307–6312; 1993.
ABBREVIATIONS
EPC—egg phosphatidylcholine EPR—electron paramagnetic resonance GC—gas liquid chromatography LPO—lipid peroxidation PC—phosphatidyl choline PUFA—polyunsaturated fatty acids RT—room temperature SUV—small unilamellar vesicles Tempol — 4-hydroxy-2,2,6,6-tetramethylpiperidine-1oxyl Tempo—2,2,6,6-tetramethylpiperidine-1-oxyl Tempol-H — 4-hydroxy-2,2,6,6-tetramethyl-1hydroxypiperidine Tempo-H—2,2,6,6-tetramethyl-1-hydroxypiperidine
/ 2b27 2426 Mp 1174 Wednesday Mar 12 04:18 PM EL–FRB (vol.22#6 ’97) 2426