Original Contribution

THE EFFECTS OF DERIVATIVES OF THE NITROXIDE TEMPOL ONUVA-MEDIATED IN VITRO LIPID AND PROTEIN OXIDATION

ELISABETTA DAMIANI , RICCARDO CASTAGNA, and LUCEDIO GRECI

Dipartimento di Scienze dei Materiali e della Terra, Universita` di Ancona, Ancona, Italy

(Received 27 December 2001;Revised 12 March 2002;Accepted 12 April 2002)

Abstract—Derivatives of tetramethylpiperidines are extensively employed in polymers to prevent photooxidation, andtheir stabilizing effect is attributed to the activity of the nitroxide radical derived from the parent amine. In this study,we examined the photoprotective effect of a commercial polymer photostabilizer, HALS-1, its corresponding nitroxide,bis(2,2,6,6-tetramethyl-piperidine-1-oxyl-4-yl)sebacate (TINO), and two derivatives of the piperidine nitroxide TEM-POL, 2,2,6,6-tetramethyl-piperidin-4-acetyloxy-1-oxyl (TEMP2) and 2,2,6,6-tetramethyl-piperidin-4-octanoyloxy-1-oxyl (TEMP8) synthesized by us, in liposomes exposed to ultraviolet A (UVA) radiation. For comparison, theUVA-absorber, 4-tert-butyl-4�-methoxydibenzoylmethane (Parsol 1789) used in many suncream formulations, was alsoincluded. The nitroxide TINO resulted extremely efficient at inhibiting aldehydic breakdown products deriving from 30min exposure of liposomes to UVA and the protection was dose-dependent (10–100�M). The corresponding amineHALS-1 was the least efficient while protection increased in the order: TEMP2� Parsol 1789� TEMP 8. HALS-1,TINO, and the two TEMPOL derivatives were also tested in a simple protein system consisting of bovine serum albumin(BSA) exposed to UVA. In this case, these compounds did not inhibit nor enhance UVA-mediated protein carbonylformation in BSA. The differences in protection between the compounds are discussed in relation to their chemicalreactivity, UVA-absorbing capacities, and their molecular structure. Overall, the results obtained envisage the potentialuse of nitroxide compounds as topical antioxidants. © 2002 Elsevier Science Inc.

Keywords—Piperidine nitroxides, Antioxidants, Parsol 1789, UVA damage, Liposomes, Proteins, Hindered amine lightstabilizers, Free radicals

INTRODUCTION

Ultraviolet radiation has long been known to be respon-sible for the degradation of polymeric systems. Polypro-pylene, for example, is particularly prone to solar radia-tion-initiated deterioration and, when not properlystabilized, has a useful lifetime of only a few months [1].Because of the widespread practical use of various goodsfabricated from polymers, stabilizers are added to tech-nical polymers to overcome the problems associated withultraviolet-mediated damage [2,3]. From this point ofview, the discovery and commercialization of a group ofcompounds known as hindered amine light stabilizers(HALS) has certainly been the most significant recentachievement in the field of polymer stabilization [4–6].

They are principally applied to a number of polymerssubjected to photooxidation, of which the most importantare polyolefins, followed by styrenics, polyurethanes,and polyamides. Recently, they have also been used inthe photostabilization of industrial coatings, especially ofthose used for automotive top coats [7]. Most of thesecompounds are derivatives of 2,2,6,6-tetramethylpiperi-dine (TEMPAMINE) (Fig. 1) and their stabilization ef-fect is attributed to the activity of the nitroxide radicalderived from the parent amine [4–6]. Our research grouphas for several years been studying the chemical reac-tivity [8–16] and antioxidant activity [17–22] in biolog-ical systems of different kinds of nitroxide radicals in-cluding those of the tetramethylpiperidine type. Hence,we thought it of interest to evaluate the photoprotectiveeffect of HALS, discovered previously in polymeric sys-tems, in biological systems. For this purpose, the widelyused, low molecular weight HALS-1 (trade name: TINU-VIN 770) along with its parent nitroxide, TINO and

Address correspondence to: Elisabetta Damiani, Dipartimento diScienze dei Materiali e della Terra, Universita` di Ancona, Via BrecceBianche, I-60131, Ancona, Italy; Tel:�39-(071)2204416; Fax:�39-(071)2204714; E-Mail: liz@unian.it.

Free Radical Biology & Medicine, Vol. 33, No. 1, pp. 128–136, 2002Copyright © 2002 Elsevier Science Inc.Printed in the USA. All rights reserved

0891-5849/02/$–see front matter

PII S0891-5849(02)00876-6

128

derivatives of the piperidine nitroxide TEMPOL (4-hy-droxy-2,2,6,6-tetramethylpiperidine-1-oxyl), TEMP2 andTEMP8 (Fig. 1) were tested in a liposomal system exposedto ultraviolet A (UVA) irradiation. In addition, the effect ofa UVA-blocking agent, 4-tert-butyl-4�-methoxydibenzoyl-methane (Parsol 1789) used in many suncream formula-tions [23,24], was also included in the study for compara-tive reasons. Furthermore, HALS-1, TINO, and theTEMPOL derivatives were also tested in a simple modelprotein system consisting of bovine serum albumin (BSA)exposed to UVA irradiation.

MATERIALS AND METHODS

Materials

BSA (Fraction V, A-6003) and L-�-phosphatidylcho-line [P2772 (Type XI-E)] were purchased from SigmaChemical Co. (Milan, Italy), bis(2,2,6,6-tetramethyl-pi-peridine-4-yl)sebacate (HALS-1) and its correspondingnitroxide, bis(2,2,6,6-tetramethyl-piperidine-1-oxyl-4-yl)sebacate (TINO) were a gift from Ciba SpecialtyChemicals (Basel, Switzerland), while all other reagentsand solvents were purchased from Aldrich Chemical Co.(Milwaukee, WI, USA). Parsol 1789 was obtained in theform of Eusolex 9020 from Merck (Darmstadt, Germa-ny). The nitroxides TEMP2 and TEMP8 (Fig. 1) weresynthesized according to the method previously de-scribed by Rosantsev, by reacting 4-hydroxy-TEMPOwith the appropriate acyl chloride and triethylamine [25].

In particular, acetyl chloride was used for the preparationof TEMP2 while octanoyl chloride was used for TEMP8.4-Hydroxy-TEMPO (0.5 mmoles) were reacted with 2mmoles of the appropriate chloride and 2 mmoles of trieth-ylamine in 20 ml toluene under magnetic stirring and atroom temperature. After 1 h the reaction was complete. Thereaction mixture was washed several times with distilledwater, extracted with 40 ml dichloromethane, dried overanhydrous sodium sulphate and concentrated. In both cases,the nitroxides were pale orange oils and used as obtained.The identity and purity of the compounds were checked bythin-layer chromatography, by mass spectroscopy on aCarlo Erba QMD 1000 spectrometer (Milan, Italy) in EI�

mode and by electron spin resonance spectroscopy on anESR spectrometer (E4; Varian, Sunnyvale, CA, USA). TheESR spectra of the nitroxides recorded in chloroformshowed the three typical bands of the nitrogens and in bothcases aN (NO•) � 15.83, g-factor � 2.00610. Their massspectra showed the following: TEMP2, calculated forC11H20NO3, 214.28, found: m/e (relative intensity), 214(M�, 28), 200 (21); TEMP8, calculated for C17H32NO3,298.45, found: m/e (relative intensity), 298 (M�, 42), 284(15).

Peroxidation of multilamellar PC liposomes inducedby UVA irradiation

PC multilamellar liposomes were prepared as follows.The desired amount of egg PC in chloroform was added

Fig. 1. Chemical structures of: 2,2,6,6-tetramethyl-piperidine (TEMPAMINE); bis(2,2,6,6-tetramethyl-piperidine-4-yl)sebacate(HALS-1); bis(2,2,6,6-tetramethyl-piperidin-1-oxyl-4-yl)sebacate (TINO); 2,2,6,6-tetramethyl-piperidin-4-acetyloxy-1-oxyl (TEMP2);2,2,6,6-tetramethyl-piperidin-4-octanoyloxy-1-oxyl (TEMP8); 4-tert-butyl-4�-methoxydibenzoylmethane (Parsol 1789).

129Nitroxides protect lipids from UVA damage

to a test tube kept in an ice bath and the solvent wasthoroughly removed under a stream of nitrogen. Whencompounds were to be tested, the desired amount of anacetonitrile solution of the compound was introducedinto another test tube, and after solvent evaporation, eggPC was added and subjected to the same procedure asdescribed above. The lipid films prepared were eachdispersed in 2 ml of 5 mM phosphate buffer, 0.9% NaCl,0.1 mM ethylenediaminetetraacetic acid (EDTA), pH7.4, and vortexed for 10 min until a white, homogeneous,opalescent suspension was obtained. The final concen-tration of the resulting multilamellar dispersion was 3.5mM. The liposomes were then transferred into 3 mlquartz cuvettes, closed and exposed to UVA irradiationat a distance of 2.5 cm from the light source for thedesired amount of time. As UVA irradiating source, aPhilips Original Home Solarium (model HB 406/A; Phil-ips, Groningen, Holland) equipped with a 400 wattozone-free Philips HPA lamp, UV type 3 was used. Thefluence rate of the light between 300 and 400 nm is 0.75mW/cm2. It was always prerun for 15 min to allow theoutput to stabilize.

The extent of lipid peroxidation was assessed using amodified method of the thiobarbituric acid (TBA) assay[26]. The concentration of thiobarbituric acid reactivesubstances (TBARS) was measured and in most casesexpressed as the percentage inhibition of TBARS forma-tion. At the end of UVA exposure, 2 ml of TBA-TCA-HCl (0.375% w/v TBA, 15% w/v TCA, 0.2 M HCl) wasadded to 0.6 ml of sample containing butylated hydroxy-toluene (BHT) 0.3 mM to prevent possible peroxidationof liposomes during the TBA assay. The samples wereheated for 20 min at 95°C followed by cooling andcentrifugation. The absorbance of the pink chromophoreof the supernatant was measured at 535 nm.

In some experiments, the compounds to be testedwere not incorporated inside the liposomes but weredissolved in propylene glycol and placed in a 2 mm widequartz cuvette. This thin cuvette was placed right in frontof a standard quartz cuvette containing the liposomalsuspension and the remaining three sides of the standardcuvette were blacked off with isolating tape. The lipo-somal suspension with the thin propylene glycol filmplaced in front of it, was exposed to UVA irradiation andsubjected to the TBA assay as described above.

Protein oxidation induced by UVA irradiation

The protein samples were prepared by dissolving 3mg/ml of BSA in 50 mM potassium phosphate buffer,0.1 mM EDTA, pH 7.4. The compounds to be testedwere added to the protein as methanol solutions (5 �l) atthe final concentration desired and the mixture was vor-texed for thorough incorporation. The samples were then

transferred into 3 ml quartz cuvettes, closed, and exposedto UVA irradiation at a distance of 2.5 cm from the lightsource. The degree of protein oxidation was monitoredby the method of Levine et al., which uses the reaction of2,4-dinitrophenylhydrazine (DNPH) with the carbonylgroups of oxidized proteins [27]. Briefly, 0.5 ml of 20mM DNPH in 2.5 M HCl was added to 0.5 ml of eachsample; blank samples lacked DNPH. Following 1 h ofincubation at room temperature with continuous shaking,the protein was precipitated by addition of 2 ml 20%TCA and centrifuged at 3000 � g for 10 min. Theprotein was washed twice with ethanol:ethylacetate (1:1)and dissolved in 1 ml of 6 M guanidine HCl, pH 6.5.Protein carbonyls were then read at 370 nm and evalu-ated using a molar absorption coefficient of 22,000M�1cm�1.

Optical absorption spectra

The absorption spectrum of all the compounds studiedwere performed in propylene glycol at a final concentra-tion of 30 �M between the wavelength range 190–500nm.

Appropriate controls were carried out throughout allthe experiments described above and the results reportedare an average of at least three independent experimentseach performed in duplicate. Statistical comparisonswere performed between: (i) samples not irradiated andsamples UVA irradiated for 10, 20, and 30 min and (ii)samples with TINO and samples with the other com-pounds tested, after 30 min irradiation, using the Stu-dent’s t-test. Differences were regarded as statisticallysignificant when p values were � 0.5 (*), p � .01 (**),and p � .001 (***).

RESULTS

The main aim of this study was to determine thephotoprotective effects of derivatives of the nitroxideTEMPOL involved in polymer stabilization, on PC lipo-somes when exposed to UVA. For comparison, the UVAchemical sunscreen, Parsol 1789, was also employed. Inaddition, the oxidative status of BSA after exposure toUVA in the presence of the TEMPOL derivatives wasalso studied.

It has previously been shown that UVA causes lipidperoxidation in liposomal membranes [28,29] and a pop-ular method for determining the extent of this process isto evaluate the percentage inhibition of aldehydic break-down products (TBARS) produced during lipid peroxi-dation, using the TBA assay [30]. Figure 2 shows thelevel of TBARS measured in liposomal suspensions be-fore and after 10, 20, and 30 min exposure to UVA

130 E. DAMIANI et al.

radiation. It can be clearly observed that after 30 minirradiation, there is almost a 6-fold increase in the levelof oxidation, and that the extent of oxidation is correlatedwith the amount of time the samples are exposed toUVA.

Figure 3 shows the results obtained when multilamel-lar liposomes were exposed to 30 min UVA irradiation,in the presence of different concentrations of the nitrox-ide TINO and Parsol 1789. It can be clearly observed thatboth at 100 �M and 50 �M, TINO is able to completelyinhibit lipid peroxidation induced by UVA exposure andas can be observed from the figure, this protective effectis concentration-dependent. At 10 �M, TINO is stillcapable of offering a good degree of protection (40%) to

liposomes exposed to UVA. For comparison, the datausing Parsol 1789 incorporated into liposomes and ex-posed to UVA is shown in the same figure. In theseexperimental conditions, Parsol 1789 was able to inhibitlipid peroxidation to some extent, although far less withrespect to TINO. Only 45% inhibition was noted at thehighest concentration used, 100 �M, while no inhibitionwas observed at the lowest concentration, 10 �M.

Experiments were also performed in which the twocompounds, TINO and Parsol 1789 were not incorpo-rated inside the liposomes, but were both separatelydissolved in propylene glycol at a final concentration of1 mM, placed as a 2 mm thick film in front of a liposomalsuspension and exposed to UVA for 30 min (see Mate-rials and Methods). The TBA assay was then performed.This experiment was carried out because it was of inter-est to see how the two compounds behaved when appliedexternally to the liposomal suspension (i.e., not in con-tact with the experimental system) instead of being in-corporated inside it. As expected, the presence of the filmof UVA-absorber placed between the liposomes andUVA rays, totally inhibited lipid peroxidation, whereaswhen TINO was used, lipid peroxidation took place(results not shown).

Figure 4 shows the percentage inhibition of TBARSformation in liposomes exposed to UVA irradiation for30 min, in the presence of the ester derivatives of thenitroxide TEMPOL, TEMP2, and TEMP8 at 100 �Mfinal concentration. The hindered amine light stabilizercorresponding to TINO, HALS-1, was also tested in theabove mentioned liposomal system at the same concen-tration and the results reported are compared to the dataobtained with TINO at the same concentration. As can beobserved, the monofunctional nitroxide TEMP8, which

Fig. 2. Concentration of TBARS determined in PC multilamellar lipo-somes (3.5 mM) in 5 mM phosphate-buffered saline, 0.1 mM EDTA,pH 7.4 after different UVA exposure times as described in Materialsand Methods (n � 3).

Fig. 3. Percentage inhibition of TBARS formation by different con-centrations of TINO and Parsol 1789 incorporated into PC multilamel-lar liposomes (3.5 mM) in 5 mM phosphate-buffered saline, 0.1 mMEDTA, pH 7.4 and exposed to UVA irradiation for 30 min. Forexperimental conditions see Materials and Methods (n � 3).

Fig. 4. Percentage inhibition of TBARS formation by 100 �M deriv-atives of TEMPOL incorporated into PC multilamellar liposomes (3.5mM) in 5 mM phosphate-buffered saline, 0.1 mM EDTA, pH 7.4 andexposed to UVA irradiation for 30 min. For experimental conditionssee Materials and Methods (n � 3).

131Nitroxides protect lipids from UVA damage

has the same carbon side chain as its correspondingbifunctional nitroxide TINO (Fig. 1), offers a good de-gree of protection (80%). Instead, TEMP2, which issimilar to TEMP8 except that it lacks the carbon sidechain, protects very little (20%). On comparing these tworesults with TINO, it is clear that the bifunctional nitrox-ide acts as a better photoprotector in this experimentalsystem, followed by the mono-nitroxide with longer sidechain. The bifunctional amine, HALS-1, proved to beonly slightly effective; in fact only 15% inhibition ofTBARS formation was observed.

To extend our investigation on the photoprotectiveeffects of the TEMPOL derivatives, HALS-1 and TINO,these were tested in a protein system consisting of BSAexposed to UVA irradiation. The degree of oxidativemodification was measured by monitoring the carbonylcontent—an index of oxidative damage to proteins [27].The results obtained (not shown) demonstrated that therewas an 8-fold increase in the carbonyl content of BSAafter exposure to UVA and that all the compounds at 100�M concentration did not significantly inhibit nor en-hance this UVA-induced protein oxidation.

DISCUSSION

Until not long ago, UVA radiation was perceived tobe beneficial to the skin, and prevention against acutesunburn mainly concentrated on protection against theharmful effects of exposure to UVB rays [31]. However,upon deeper investigation, it was discovered that UVA ismost cytotoxic to human skin cells. These rays penetratedeeper into the dermis, where it provokes dermal con-nective tissue alterations associated with photoaging andmany other subchronic to chronic skin disorders [32,33].In fact, UVA has been shown to induce lipid peroxida-tion in liposomal, micellar, and natural systems [28,29],and our results reported in Fig. 2 confirm this finding.

Membrane leakage [34], DNA strand breaks [35], pyrim-idine dimers and adducts [36] as well as damage toproteins and enzymes [37] consequent to UVA exposure,have also been reported. For these reasons, much re-search interest exists in the field of photoprotection forfinding ways of reducing sun exposure damage. In poly-mers, this problem has been largely overcome by the useof photostabilizers, the most popular of which are HALS[4–6].

The principal feature of HALS activity in protectionagainst photo-oxidation includes the ability of the parentamine and its transformation products to interfere withchain oxidation by scavenging radicals or by disarmingthe initiators, although there is still enormous theoreticalinterest in deciphering their antioxidant mechanism [38].Nonetheless, the photo-antioxidant effect of HALS hascertainly been connected with the efficiency of the de-rived nitroxides [4–6] especially those having a shorteralkylidene bridge connecting the two piperidine nuclei,such as HALS-1 [39]. An oversimplification of the prob-able steps involved in the antioxidant mechanism of theamines and the nitroxides are shown in Fig. 5. Briefly,during polymer (RH) photo-oxidation, oxygen-centeredradicals, such as peroxyl radicals (ROO•) are formed.Hydrogen abstraction by these peroxyls on HALS(�N-H) leads to the formation of the aminyl radical(�N•) which reacts with oxygen generating the parentnitroxide (�N-O•). The nitroxide can then deactivatealkyl radicals (R•), a necessary condition for efficientstabilization of polymers in the earliest steps of oxidativedegradation, leading to the formation of the respectiveO-alkylhydroxylamine (�NOR). Alternatively, nitrox-ides may abstract hydrogen leading to hydroxylamines(�NOH). These two latter species can both act as chain-breaking antioxidants, and in so doing, the parent nitrox-ide is regenerated. However, detailed accounts of all

Fig. 5. Pathways for inhibition of oxidative processes by nitroxides and their corresponding amines.

132 E. DAMIANI et al.

these steps have been reported elsewhere and will not bedealt with here [4–6,38,39].

In this study we now report that the bifunctionalnitroxide TINO, deriving from a hindered amine lightstabilizer, HALS-1, is also an efficient photo-antioxidantwhen incorporated in lipid systems exposed to UVAlight, whereas its parent amine was not at all as effective.The incapability of HALS-1 to efficiently protect lipo-somes against UVA-mediated damage in our experimen-tal model was quite surprising, because a good degree ofprotection was expected, considering the performance ofthese compounds as long-term photostabilizers for poly-mers. The slight protection conferred by HALS-1 maylikely arise from the deactivation of peroxy radicals asdescribed in Fig. 5. In addition, because singlet oxygen isbelieved to be the main species involved at the level ofthe formation of PC hydroperoxides mediated by UVA[37,40] and that amines of the piperidine type, likeTEMPAMINE (Fig. 1) react with singlet oxygen to givethe corresponding nitroxide [41], some of the protectionobserved by HALS-1 is thought to occur in this way. Atpresent, we have no explanation for the modest photo-protection of HALS-1 reported in this study, although itcan safely be said that the results obtained in one system,i.e., polymers, may not necessarily translate into a sim-ilar degree of protection in another totally different sys-tem, in our case, lipids. However, the encouraging resultobtained with TINO confirms what has previously beenobserved in polymer systems: that the nitroxide groupplays an important role in deactivating the species gen-erated during UVA exposure of PC liposomes under theexperimental conditions used in this work. Hence, theonly explanation for the protection conferred by thenitroxide is due to scavenging of reactive radical speciessuch as alkyls involved during the chain oxidation thatgoverns lipid peroxidation. This is a well-studied reac-tion investigated several years ago by Ingold and co-workers, who showed that nitroxide radicals react withalkyl radicals at an almost diffusion controlled rate [42].

Interesting results were observed when the efficiencyof TINO was compared to that of a commercially usedUVA-blocker, Parsol 1789 (Fig. 1), added in many sun-cream formulations. This UVA sun filter, which absorbsultraviolet light thereby reducing sunburn, has beenshown in previous studies to generate free radicals uponillumination [43,44] and that these are most probablyresponsible for the enhanced protein damage [18] andplasmid DNA strand breaks [19] that has been observedin those experimental systems. The results reported inFig. 3 show that, contrary to what was found with plas-mid DNA and BSA, Parsol 1789 protects UVA-mediatedlipid peroxidation, though to a much lesser degree com-pared to TINO. This result is comforting considering thewidespread use of sunfilters in many cosmetic products,

and in view of the fact that a bilipid layer is the firstbarrier that external agents encounter when coming intocontact with cells. The reasons for this protection, de-spite the formation of free radicals, is most probably dueto its capability to absorb some of the UVA radiationresponsible for the lipid damage observed, but that at theconcentrations used, this was not sufficient to totallyinhibit the oxidative process as well as TINO does.However, when a propylene glycol solution of Parsol1789 was placed in a 2 mm thick, flat quartz cuvette andthis was positioned between the liposomal suspensionand the irradiating source, therefore, not in contact withthe liposomes (see Materials and Methods), total protec-tion against lipid peroxidation was observed. This obvi-ously implies that the sunfilter is an efficient UVA ab-sorber when it is present externally to the lipid systemunder study but that it is less efficient when in contactwith it. On the contrary, when TINO was added exter-nally as described above, lipid peroxidation occurred.Hence, as mentioned earlier, protection by TINO is at-tributed to its radical scavenging ability because it doesnot absorb light in the UVA region. The optical absorp-tion spectra of Parsol 1789 and TINO shown in Fig. 6confirm the above statements; Parsol 1789 shows strongabsorption between 320–400 nm (UVA range) whereasTINO at the same concentration does not absorb at all.

The interesting results obtained with TINO stimulatedus to synthesize and study some derivatives of TEMPOLin order to find further explanations, from the structuralpoint of view, for TINO’s high photoantioxidant effectobserved in liposomal systems. In particular, two deriv-atives of varying side chain length, TEMP2 and TEMP8were studied. TEMP8 is the monofunctional derivativeof TINO, whereas TEMP2 lacks the octyl side chain aswell as a nitroxide function. From the results reported in

Fig. 6. Optical absorption spectra of the compounds tested, in pro-pylene glycol, at a final concentration of 30 �M recorded at roomtemperature.

133Nitroxides protect lipids from UVA damage

Fig. 4, it is possible to observe that TEMP8 is almost justas efficient as TINO in protecting UVA-mediated lipidperoxidation. Because these two compounds differ byonly one additional nitroxide function, it may safely besaid that the 20% difference in protection observed be-tween TINO and TEMP8 may be accounted for by theextra nitroxide function that is responsible for the higherantioxidant effect. The short chain analogue TEMP2,which is devoid of side chain, is only slightly protective,most probably because it is not as lipid soluble as TINOand TEMP8. In fact, in multilamellar liposomes, UVA-radiation-dependent peroxidation occurs at sites less ac-cessible to the aqueous phase [37]. Hence, to have pro-tective effects, an antioxidant has to reside in closeproximity to the membrane interior region where lipidperoxidation chain propagation occurs.

Protection of UVA-mediated protein oxidation by theTEMPOL derivatives was not as remarkable as in thelipid system studied and the slight protections observedwere not statistically significant (results not shown). Theprotein radicals generated during photooxidation ofBSA, such as thiyl radicals and tryptophan-derived rad-icals [45], are most probably in sites or pockets lessaccessible to the nitroxide compounds or of a certainchemical nature which are not efficiently trapped by theantioxidants, hence the negligible antioxidant effect. Ofinterest however, is the observation that the compoundsadded to BSA do not lead to any increases in oxidativemodifications when illuminated, contrary to what haspreviously been observed with Parsol 1789 [18].

In conclusion, by using UVA irradiation, where theTEMPOL derivatives TINO, TEMP8 and TEMP2 do notabsorb (Fig. 6), we have demonstrated that with theexception of TEMP2, the other two compounds dramat-ically protect against UVA-enhanced lipid damage andthis protection is correlated with the chemical structure.Secondly, when TINO was incorporated in liposomes, itproved to be a far better photo-antioxidant than theUVA-absorber, Parsol 1789. The corresponding bifunc-tional amine of TINO, HALS-1 used as a light stabilizerin polymers, was only slightly photo-protective in thisbiological system. Because the outstanding antioxidantactivity of nitroxide radicals in many diverse simple andcomplex biological systems is now notoriously knownand attributed to their radical scavenging ability, super-oxide dismutase mimicry and oxidation of transitionmetals thereby reducing Fenton-type reactions [46–49],the results reported here further corroborate the possibleemployment of these interesting compounds to preventdisorders such as those related to photoaging. In fact, theresults reported here are in accordance with the recentfindings of Bernstein et al., which have shown that thenitroxide TEMPOL provided over 50% protectionagainst UVB- and over 70% protection against UVA-

induced damage in a transgenic murine fibroblast culturemodel of cutaneous photoaging [50]. Protection againstUV damage by nitroxide radicals in combination with abenzotriazole-based UV absorber has also been observedin milled wood lignin and filter paper that undergoesyellowing after UVA exposure [51]. It is also worthmentioning that the highly lipophilic hydroxylamine de-rivative of HALS-1 quantitatively reacts with oxygen-centered radicals to give the corresponding stable nitrox-ide, and for this reason it has been successfully employedas spin probe to measure reactive oxygen species inhuman tissues [52]. From this point of view, the hydrox-ylamine derivatives of HALS-1 and TEMP8 might pro-vide protection to liposomes exposed to UVA as well,especially because previous findings have shown thatN-hydroxy derivatives protect biological molecules justas well as their parent nitroxide [12,21].

Because UVA irradiation causes depletion of antioxi-dants and oxidative damage in models of human skin[53] and that both oral supplementation or topical appli-cation with antioxidants such as vitamin E and ascorbicacid reduces the sunburn reaction [54,55], it could likelybe envisaged that nitroxide radicals may too be includedin formulations aimed at reducing this damage. It hasalready been reported that topical application of nitrox-ide radicals protects against radiation-induced alopecia[56] and that nitroxides are relatively non-toxic to skin[57], therefore the above considerations on their possibleuse as topical antioxidants may not be that far from thereality.

Acknowledgements — The authors wish to thank the Italian MURST(Ministero dell’Universita e della Ricerca Scientifica e Tecnologica)and the University of Ancona for financial support.

REFERENCES

[1] Allen, N. S.; Chirinos-Padron, A.; Henman, T. J. The photosta-bilisation of polypropylene: a review. Polym. Degrad. Stab. 13:31–38; 1985.

[2] Faucitano, A.; Buttafava, A.; Camino, G.; Greci, L. Photo-oxida-tion and stabilization of polymers. Trends Polym. Sci. 4:92–98;1996.

[3] Gugumus, F. Photooxidation of polymers and its inhibition. In:Pospisil, J.; Klemchuk, P. P., eds. Oxidation inhibition in organicmaterials, vol. II. Boca Raton, FL: CRC Press, Inc.; 1990:29–161.

[4] Sedlar, J. Hindered amines as photostabilizers. In: Pospisil, J.;Klemchuk, P. P., eds. Oxidation inhibition in organic materials,vol. II. Boca Raton, FL: CRC Press, Inc.; 1990:2–28

[5] Rosantsev, E. G.; Kagan, E. S.; Sholle, V. D.; Ivanov, V. B.;Smirnov, V. A. Discovery, chemistry and application of hinderedamines. In: Klemchuk, P. P., ed. Polymer stabilization and deg-radation. Washington, DC: American Chemical Society; 1985:11–35.

[6] Muller, H. K. A decade of hindered amine light stabilizers. In:Klemchuk, P. P., ed. Polymer stabilization and degradation.Washington, DC: American Chemical Society; 1985:55–68.

[7] Bauer, D. R.; Herlock, J. L.; Mielewski, D. F. Photostabilizationand photodegradation in organic coatings containing hindered

134 E. DAMIANI et al.

amine light stabilizers VII. HALS effectiveness in acrylic mel-amine coatings having different free radical formation rates.Polym. Degrad. Stab. 36:9–15; 1992.

[8] Carloni, P.; Damiani, E.; Iacussi, M.; Greci, L.; Stipa, P.; Rizzoli,C.; Sgarabotto, P. Unexpected deoxygenation of 2,2,6,6,-tetram-ethylpiperidine-1-oxyl (TEMPO) by thiyl radicals through theformation of arylsulphinyl radicals. Tetrahedron52:12445–12452; 1995.

[9] Carloni, P.; Damiani, E.; Greci, L.; Stipa, P.; Marrosu, G.;Petrucci, R.; Trazza, A. Chemical and electrochemical study onthe interactions of aminoxyls with superoxide anion. Tetrahedron51:11257–11264; 1996.

[10] Stipa, P.; Carloni, P.; Greci, L.; Damiani, E. Synthesis and ther-mal stability of alkoxyamines. Polym. Degrad. Stab. 55:323–327;1997.

[11] Cardellini, L.; Carloni, P.; Greci, L.; Stipa, P.; Faucitano, A.Homolytic substitution in indolinone nitroxide radicals. Part 5.Reactions with tert-butylperoxy radicals. Gazz. Chim. Ital. 119:621–625; 1989.

[12] Noguchi, N.; Damiani, E.; Greci, L.; Niki, E. Action of quinolinicand indolinonic aminoxyls as radical-scavenging antioxidants.Chem. Phys. Lipids 99:11–19; 1999.

[13] Greci, L. Homolytic substitution in indolinone nitroxide radicals.Part III. Reactions with terbutoxy and methyl radicals. Tetrahe-dron 38:2435–2439; 1982.

[14] Damiani, E.; Carloni, P.; Iacussi, M.; Stipa, P.; Greci, L. Reac-tivity of sulphur-centered radicals with indolinonic and quinolinicaminoxyls. Eur. J. Org. Chem. 1:2405–2412; 1999.

[15] Mar’ in, A.; Damiani, E.; Canestrari, S.; Dubs, P.; Greci, L.Interaction between trialkyl phospites and aminoxyl radicals: amodel study for polymer stabilization. J. Chem. Soc. PerkinTrans. 2:1363–1367; 1999.

[16] Damiani, E.; Greci, L.; Rizzoli, C. Reaction of indolinonic ami-noxyls with nitric oxide. J. Chem. Soc. Perkin Trans. 2:1139–1144; 2001.

[17] Damiani, E.; Kalinska, B.; Canapa, A.; Canestrari, S.; Wozniak,M.; Olmo, E.; Greci, L. The effects of nitroxide radicals onoxidative DNA damage. Free Radic. Biol. Med. 28:1257–1265;2000.

[18] Damiani, E.; Carloni, P.; Biondi, C.; Greci, L. Increased oxidativemodification of albumin when illuminated in vitro in the presenceof a common sunscreen ingredient: protection by nitroxide radi-cals. Free Radic. Biol. Med. 28:193–201; 2000.

[19] Damiani, E.; Greci, L.; Parsons, R.; Knowland, J. Nitroxideradicals protect DNA from damage when illuminated in vitro inthe presence of dibenzoylmethane and a common sunscreen in-gredient. Free Radic. Biol. Med. 26:809–816; 1999.

[20] Antosiewicz, J.; Popinigis, J.; Wozniak, M.; Damiani, E.; Carloni,P.; Greci, L. Effects of indolinic and quinolinic aminoxyls onprotein and lipid peroxidation of rat liver microsomes. FreeRadic. Biol. Med. 18:913–917; 1995.

[21] Damiani, E.; Paganga, G.; Greci, L.; Rice-Evans, C. Inhibition ofcopper-mediated low density lipoprotein peroxidation by quino-line and indolinone nitroxide radicals. Biochem. Pharmacol. 48:1155–1161; 1994.

[22] Antosiewicz, J.; Bertoli, E.; Damiani, E.; Greci, L.; Popinigis, J.;Przybylski, S.; Tanfani, F.; Wozniak, M. Indolinic and quinolinicaminoxyls as protectants against oxidative stress. Free Radic.Biol. Med. 15:203–208; 1993.

[23] De Polo, K. F. Dibenzoylmethane derivative and its use in sun-screening compositions. Chem. Abs. 93:185967g; 1980.

[24] Deflandre, A.; Lang, G. Photostability assessment of sunscreens.Benzylidene camphor and dibenzoylmethane derivatives. Int. J.Cosmet. Sci. 10:53–62; 1988.

[25] Rozantsev, E. G.; Suskina, V. I. Synthesizing new stable freeradicals in the 2,2,6,6-tetramethylpiperidin-1-oxyl series. Chem.Abs. 71:49730t; 1969.

[26] Buege, J.; Aust, S. D. Microsomal lipid peroxidation. MethodsEnzymol. 52:302–310; 1978.

[27] Levine, R. L.; Garland, K.; Oliver, C. N.; Amici, A.; Climent, I.;Lenz, A. G.; Ahn, B. W.; Shaltiel, S.; Stadtaman, E. R. Determi-

nation of carbonyl content in oxidatively modified proteins. Meth-ods Enzymol. 186:464–478; 1990.

[28] Bose, B.; Chatterjee, S. N. Correlation between UVA-inducedchanges in microviscosity, permeability and malondialdehydeformation in liposomal membrane. J. Photochem. Photobiol. BBiol. 28:149–153; 1995.

[29] Bose, B.; Agarwal, S.; Chatterjee, S. N. UV-A induced lipidperoxidation in liposomal membrane. Radiat. Enivron. Biophys.28:59–65; 1989.

[30] Esterbauer, H.; Cheeseman, K. H. Determination of aldehydiclipid peroxidation products: malondialdehyde and 4 hydroxynon-enal. Methods Enzymol. 186:407–421; 1990.

[31] De Gruijl, F. R. Photocarcinogenesis: UVA vs UVB. MethodsEnzymol. 319:359–366; 2000.

[32] Schaefer, H.; Chardon, A.; Moyal, D. Photoprotection of skinagainst ultraviolet A damage. Methods Enzymol. 319:445–465;2000.

[33] Council on Scientific Affairs. Harmful effects of ultraviolet radi-ation. J. Am. Med. Assoc. 262:380–384; 1989.

[34] Bose, B.; Agarwal, S.; Chatterjee, S. N. Membrane lipid peroxi-dation by UV-A: mechanism and implications. Biotech. Appl.Biochem. 12:557–561; 1990.

[35] Dissanayake, N. S.; Greenoak, G. E.; Mason, R. S. Effects ofultraviolet irradiation on human skin-derived epidermal cells invitro. J. Cell. Physiol. 157:119–127; 1993.

[36] Kielbassa, C.; Epe, B. DNA damage induced by ultraviolet andvisible light and its wavelength dependence. Methods Enzymol.319:436–445; 2000.

[37] Vile, G. F.; Tyrrell, R. M. UVA radiation-induced oxidativedamage to lipids and proteins in vitro and in human skin fibro-blasts is dependent on iron and singlet oxygen. Free Radic. Biol.Med. 18:721–730; 1995.

[38] Gugumus, F. Current trends in mode of action of hindered aminelight stabilizers. Polym. Degrad. Stab. 40:167–215; 1993.

[39] Pospisil, J. Aromatic and heterocyclic amines in polymer stabili-zation. In: Dusek, K., ed. Advances in polymer science. Berlin:Springer Verlag; 1995:89–189.

[40] Valenzeno, D. P. Photomodification of biological membraneswith an emphasis on singlet oxygen mechanisms. Photochem.Photobiol. 46:160–174; 1987.

[41] Hideg, E.; Vass, I.; Kalai, T.; Hideg, K. Singlet oxygen detectionwith sterically hindered amine derivatives in plants under lightstress. Methods Enzymol. 319:77–85; 2000.

[42] Chateauneuf, J.; Lusztyk, J.; Ingold, K. U. Absolute rate constantsfor the reactions of some carbon-centered radicals with 2,2,6,6-tetramethylpiperidine-N-oxyl. J. Org. Chem. 53:1629–1632;1988.

[43] Roscher, N. M.; Lindemann, M. K. O.; Kong, S. B.; Cho, C. G.;Jiang, P. Photodecomposition of several compounds commonlyused as sunscreen agents. J. Photochem. Photobiol. A Chem.80:417–421; 1994.

[44] Schwack, W.; Rudolph, T. Photochemistry of dibenzoyl methaneUVA filters. Part 1. J. Photochem. Photobiol. 28:229–234; 1995.

[45] Silvester, J. A.; Timmins, G. S.; Davies, M. J. Photodynamicallygenerated bovine serum albumin radicals; evidence for damagetransfer and oxidation at cysteine and tryptophan residues. FreeRadic. Biol. Med. 24:754–766; 1998.

[46] Mitchell, J. B.; Krishna, M. C.; Kuppusamy, P.; Cook, J. A.;Russo, A. Protection against oxidative stress by nitroxides. Exp.Biol. Med. 226:620–621; 2001.

[47] Mitchell, J. B.; Krishna, M. C.; Samuni, A. Nitroxides as protec-tors against oxidative stress. In: Colton, C.; Gilbert, D., eds.Reactive oxygen species in biological systems. New York: PlenumPress; 1999:293–313.

[48] Samuni, A.; Krishna, M. C. Antioxidant properties of nitroxidesand nitroxide SOD mimics. In: Packer, L.; Cadenas, E., eds.Handbook of synthetic antioxidants. New York: Marcel DekkerInc.; 1997:351–373.

[49] Kocherginsky, N.; Swartz, H. M. Nitroxide spin labels; reactionsin biology and chemistry. Boca Raton, FL: CRC Press; 1995.

[50] Bernstein, E. F.; Kong, S. K.; Brown, D. B.; Kwak, B. C.;

135Nitroxides protect lipids from UVA damage

Takeuchi, T.; Gasparro, F. P.; Uitto, J. The nitroxide Tempolaffords protection against ultraviolet radiation in a transgenicmurine fibroblast culture model of cutaneous photoaging. Exp.Dermatol. 10:55–61; 2001.

[51] Beaton, C. R.; Argyropoulos, D. S. Photostabilizing milled woodlignin with benzotriazoles and hindered nitroxides. Photochem.Photobiol. 73:605–610; 2001.

[52] Valgimigli, L.; Pedulli, G. F.; Paolini, M. Measurement of oxi-dative stress by EPR radical-probe technique. Free Radic. Biol.Med. 31:708–716; 2001.

[53] Podda, M.; Traver, M. G.; Weber, C.; Yan, L.-J.; Packer, L.UV-irradiation depletes antioxidants and causes oxidative damagein a model of human skin. Free Radic. Biol. Med. 24:55–65;1998.

[54] Fuchs, J.; Kern, H. Modulation of UV-light induced skin inflam-mation by D-alpha-tocopherol and L-ascorbic acid: a clinicalstudy using solar simulated radiation. Free Radic. Biol. Med.25:1006–1012; 1998.

[55] Stab, F.; Wolber, R.; Blatt, T.; Keyhani, R.; Sauermann, G.Topically applied antioxidants in skin protection. Methods Enzy-mol. 319:465–478; 2000.

[56] Cuscela, D.; Coffin, D.; Lupton, G. P.; Cook, J. A.; Krishna,M. C.; Bonner, R. F.; Mitchell, J. B. Protection from radiation-induced alopecia with topical application of nitroxides: fraction-ated studies. Cancer J. Sci. Am. 2:273–278; 1998.

[57] Fuchs, J.; Croth, N.; Herrling, T. Cutaneous tolerance to nitroxidefree radicals in human skin. Free Radic. Biol. Med. 24:643–648;1998.

136 E. DAMIANI et al.