Ultraviolet A Induces Generation of Squalene Monohydroperoxide Isomers in Human Sebum and Skin Surface Lipids In Vitro and In Vivo

At the outermost surface of human skin, skin surface lipids are first-line targets of solar ultraviolet radiation. Therefore, we hypothesized that ultraviolet A and ultraviolet B irradiation induce photo-oxidation of skin surface lipids. To test this, sebum samples were collected from facial skin of 17 healthy volunteers, weighed, and immediately irradiated with either ultraviolet B or ultraviolet A. Squalene, the major sebum lipid, as well as photo-oxidation products were identified in sebum lipid extracts by high-performance liquid chromatography analysis. Upon ultraviolet A exposures squalene was depleted in a concentration-dependent manner, whereas an unidentified sebum lipid photo-oxidation product was detected. Using high-performance thin layer chromatography, high-performance liquid chromatography, atmospheric pressure chemical ionization mass spectrometry, and nuclear magnetic resonance, unidentified sebum lipid photo-oxidation product was identified as a mixture of squalene monohydroperoxide isomers. Squalene monohydroperoxide isomers purified from sebum were identical with squalene monohydroperoxide isomers synthesized by preparative photo-oxidation of squalene. Squalene monohydroperoxide isomers were formed even after small suberythematogenic doses of ultraviolet A (5 J per cm2). Whereas physiologic baseline levels of squalene monohydroperoxide isomers in human skin were only slightly above detection limits, squalene monohydroperoxide isomer levels were strongly increased by suberythematogenic doses of ultraviolet A both in vitro and in vivo. High-performance liquid chromatography results could be complemented by a straightforward thin layer chromatography method for rapid screening of lipid peroxide formation in human sebum/skin surface lipids. In conclusion, specific squalene monohydroperoxide isomers were identified as highly ultraviolet A sensitive skin surface lipid breakdown products that may serve as a marker for photo-oxidative stress in vitro and in vivo. At the outermost surface of human skin, skin surface lipids are first-line targets of solar ultraviolet radiation. Therefore, we hypothesized that ultraviolet A and ultraviolet B irradiation induce photo-oxidation of skin surface lipids. To test this, sebum samples were collected from facial skin of 17 healthy volunteers, weighed, and immediately irradiated with either ultraviolet B or ultraviolet A. Squalene, the major sebum lipid, as well as photo-oxidation products were identified in sebum lipid extracts by high-performance liquid chromatography analysis. Upon ultraviolet A exposures squalene was depleted in a concentration-dependent manner, whereas an unidentified sebum lipid photo-oxidation product was detected. Using high-performance thin layer chromatography, high-performance liquid chromatography, atmospheric pressure chemical ionization mass spectrometry, and nuclear magnetic resonance, unidentified sebum lipid photo-oxidation product was identified as a mixture of squalene monohydroperoxide isomers. Squalene monohydroperoxide isomers purified from sebum were identical with squalene monohydroperoxide isomers synthesized by preparative photo-oxidation of squalene. Squalene monohydroperoxide isomers were formed even after small suberythematogenic doses of ultraviolet A (5 J per cm2). Whereas physiologic baseline levels of squalene monohydroperoxide isomers in human skin were only slightly above detection limits, squalene monohydroperoxide isomer levels were strongly increased by suberythematogenic doses of ultraviolet A both in vitro and in vivo. High-performance liquid chromatography results could be complemented by a straightforward thin layer chromatography method for rapid screening of lipid peroxide formation in human sebum/skin surface lipids. In conclusion, specific squalene monohydroperoxide isomers were identified as highly ultraviolet A sensitive skin surface lipid breakdown products that may serve as a marker for photo-oxidative stress in vitro and in vivo. atmospheric pressure chemical ionization mass spectrometry electrospray ionization mass spectrometry high performance liquid chromatography high performance thin layer chromatography N,N-dimetyl-14-phenylene diamine nuclear magnetic resonance reversed phase skin surface lipids squalene monohydroperoxides total correlation spectroscopy unidentified sebum lipid photo-oxidation product The skin serves as a biologic interface between body and environment and thus is exposed to high intensities of solar ultraviolet (UV) irradiation, the major source of cutaneous oxidative stress (Fuchs, 1992Fuchs J. Oxidative Injury in Dermatopathology. Springer-Verlag, Berlin1992Crossref Google Scholar; Gilchrest, 1995Gilchrest B.A. Photodamage. Cambridge, MA1995Google Scholar; Thiele et al., 2001Thiele J.J. Schroeter C. Hsieh S.N. Podda M. Packer L. The antioxidant network of the stratum corneum.Curr Probl Dermatol. 2001; 29: 26-42Crossref PubMed Google Scholar). Besides direct UVB-induced DNA damage, photo-oxidative skin damage is mediated by reactive oxygen species, and results in anti-oxidant depletion, differential regulation of anti-oxidant enzymes, oxidative protein damage, and lipid peroxidation (Scharffetter-Kochanek, 1997Scharffetter-Kochanek K. Photoaging of the connective tissue of skin: Its prevention and therapy.Adv Pharmacol. 1997; 38: 639-655Crossref PubMed Scopus (41) Google Scholar; Thiele et al., 1998Thiele J.J. Traber M.G. Packer L. Depletion of human stratum corneum vitamin E. An early and sensitive in vivo marker of UV-induced photo-oxidation.J Invest Dermatol. 1998; 110: 756-761Abstract Full Text Full Text PDF PubMed Scopus (154) Google Scholar, Thiele et al., 1999aThiele J.J. Hsieh S.N. Briviba K. Sies H. Protein oxidation in human stratum corneum: Susceptibility of keratins to oxidation in vitro and presence of a keratin oxidation gradient in vivo.J Invest Dermatol. 1999; 113: 335-339Abstract Full Text Full Text PDF PubMed Scopus (105) Google Scholar, Thiele et al., 2001Thiele J.J. Schroeter C. Hsieh S.N. Podda M. Packer L. The antioxidant network of the stratum corneum.Curr Probl Dermatol. 2001; 29: 26-42Crossref PubMed Google Scholar; Sander et al., 2002Sander C.S. Chang H. Salzmann S. Müller C.S.L. Ekanayake-Mudiyanselage S. Elsner P. Thiele J.J. Photoaging is associated with protein oxidation in human skin in vivo.J Invest Dermatol. 2002; 118: 618-625Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar). Lipid peroxide and lipid hydroperoxide generation occurs during oxidative damage in biologic membranes (Glenn and Tyrell, 1995Glenn V.F. Tyrell R.M. UVA radiation-induced oxidative damage to lipids and proteins in vitro and in human skin fibroblasts is dependent on iron and singlet oxygen.Free Radic Biol Med. 1995; 18: 721-730Crossref PubMed Scopus (243) Google Scholar; Girotti, 1998Girotti A.W. Lipid hydroperoxide generation, turnover, and effector action in biological systems.J Lipid Res. 1998; 39: 1529-1542Abstract Full Text Full Text PDF PubMed Google Scholar). To counteract lipid peroxidation, the skin is equipped with a network of lipophilic and hydrophilic, enzymatic, and nonenzymatic anti-oxidant systems that are closely interlinked (Thiele et al., 2000Thiele J.J. Dreher F. Packer L. Antioxidant defense systems in skin.in: Elsner P. Maibach H. Drugs Vs. Cosmetics: Cosmeceuticals? Marcel Dekker, New York2000: 145-188Google Scholar). Vitamin E was identified as the predominant anti-oxidant in the uppermost human skin layers, the stratum corneum and skin surface lipids (SSL) (Thiele et al., 2001Thiele J.J. Schroeter C. Hsieh S.N. Podda M. Packer L. The antioxidant network of the stratum corneum.Curr Probl Dermatol. 2001; 29: 26-42Crossref PubMed Google Scholar; Thiele, 2001Thiele J.J. Oxidative targets in the stratum corneum: a new basis for antioxidative strategies.Skin Pharmacol Appl Skin Physiol. 2001; 14: 87-91Crossref PubMed Google Scholar). Interestingly, vitamin E concentrations in human facial skin are highest at the skin surface layers, i.e., the stratum corneum and the overlaying SSL (Thiele et al., 1999bThiele J.J. Weber S.U. Packer L. Sebaceous gland secretion is a major physiological route of vitamin E delivery to skin.J Invest Dermatol. 1999; 113: 1006-1010Abstract Full Text Full Text PDF PubMed Scopus (171) Google Scholar). Of all the skin layers, the stratum corneum and SSL are exposed to the highest intensities of solar UV radiation (Parrish, 1983Parrish J.A. Responses of skin to visible and ultraviolet radiation.in: Goldsmith L.A. Biochemistry and Physiology of the Skin. Vol. 2. Oxford University Press, New York1983: 713-733Google Scholar). It has been demonstrated by in vivo chemiluminescence that UVA exposure strongly increases the formation of reactive oxygen species in the skin's uppermost layers (Evelson et al., 1997Evelson P. Ordóñez C.P. Llesuy S. Boveris A. Oxidative stress and in vivo chemiluminescence in mouse skin exposed to UVA radiation.J Photochem Photobiol B Biol. 1997; 38: 215-219Crossref PubMed Scopus (64) Google Scholar). Thus, SSL, which are covering the stratum corneum in sebaceous gland rich anatomical regions, such as the face, have to be considered as first-line-targets of solar UV exposure. SSL are derived from epidermal lipids as well as from sebaceous gland lipids (sebum) (Elias, 1983Elias P.M. Epidermal lipids, barrier function, and desquamation.J Invest Dermatol. 1983; 80: 44-49Abstract Full Text PDF PubMed Google Scholar; Wertz and Downing, 1991Wertz P.W. Downing D.T. Epidermal lipids.in: Goldsmith L.A. Physiology, Biochemistry, and Molecular Biology of the Skin. Oxford University Press, New York1991: 205-236Google Scholar; Downing, 1992Downing D.T. Lipid and protein structures in the permeability barrier of mammalian epidermis.J Lipid Res. 1992; 33: 301-313Abstract Full Text PDF PubMed Google Scholar). The main component of human sebum is squalene, an unsaturated lipid, which is generated in sebaceous glands (Stewart and Downing, 1991Stewart M.E. Downing T.D. Chemistry and function of mammalian sebaceous lipids.in: Elias P.M. Skin Lipids: Advances in Lipid Research. Vol. 24. Academic Press, Inc., San Diego1991: 263-301Google Scholar). Recently, we have demonstrated that human sebum contains high levels of α-tocopherol, which are secreted on to the skin surface and thus contribute to SSL (Thiele et al., 1999bThiele J.J. Weber S.U. Packer L. Sebaceous gland secretion is a major physiological route of vitamin E delivery to skin.J Invest Dermatol. 1999; 113: 1006-1010Abstract Full Text Full Text PDF PubMed Scopus (171) Google Scholar). It was postulated that the biologic role of the high concentrations of vitamin E present in SSL may serve as protection against photo-oxidation. To investigate the susceptibility of SSL to photo-oxidative stress, human sebum was collected and subjected to defined doses of UVA and UVB. This study was based on preliminary results that had revealed the accumulation of unexpectedly high amounts of an unidentified lipid photo-oxidation product in human sebum (USLPP) upon physiologic and suberythematogenic levels of UVA. 1Ekanayake Mudiyanselage S, Elsner P, Thiele JJ: Wavelength dependent UV-induced depletion of vitamin E and generation of a highly sensitive lipid photo-oxidation product in human sebum. Arch Dermatol Res 294:72, 2002 (Abstr.).1Ekanayake Mudiyanselage S, Elsner P, Thiele JJ: Wavelength dependent UV-induced depletion of vitamin E and generation of a highly sensitive lipid photo-oxidation product in human sebum. Arch Dermatol Res 294:72, 2002 (Abstr.). The goal of this study was: (i) to develop a sensitive and reliable method for detecting SSL photo-oxidation products; (ii) to identify the nature and source of USLPP; (iii) to investigate the wavelength dependency of SSL photo-oxidation; and (iv) to quantitate and compare the formation of SSL photo-oxidation products in vitro and in vivo. All chemicals and solvents were of the highest analytical or high performance liquid chromatography (HPLC) grade, unless specified otherwise. HPLC grade ethanol, methanol, benzene, toluene, and ethyl acetate were from Roth GmbH (Karlsruhe, Germany). Rose Bengal disodium salt (3′,4′,5′,6′-tetrachloro-2,4,5,7-tetraiodo fluorescein disodium salt) was from VEB Laborchemie (Apolda, Germany) and squalene was from Sigma-Aldrich Chemie (Steinheim, Germany). SSL of the face were collected from 17 healthy volunteers [nine females, eight male; average age 30.7±6.2 y (mean±SD); Fitzpatrick skin types I–IV (I/II: one, II/III: 14, III/IV: two)]. Permission was granted by the local ethic committee of the Friedrich-Schiller-University of Jena. Exclusion criteria were any history of dermatologic disorders and any current medical problems or systemic medication. The participants were not allowed to apply creams and ointments on the face or to take any oral or topical anti-oxidant supplementation 2 wk prior to and during the time of the study. Samples of sebum were collected from the forehead of the volunteers using Sebutapes® (Cuderm, Dallas, TX) as previously described (Thiele and Packer, 1999Thiele J.J. Packer L. Non-invasive measurement of alpha-tocopherol gradients in human stratum corneum by HPLC analysis of sequential tape strippings.Methods Enzymol. 1999; 300: 413-419Crossref PubMed Scopus (18) Google Scholar). Briefly, forehead skin was cleaned prior to sample acquisition using a sterile gauze ball (Gazin®, Lohmann & Rauscher International GmbH, Rengsdorf, Germany) soaked in 1 ml 70% ethanol solution. Each tape was weighed before and after sebum collection. The sebum collection time was 1 h for every tape. The average amount of sebum collected per Sebutape® in 1 h was 1.03±0.4 mg (n=202; mean±SD). Immediately after collection and weighing, sebum-enriched tapes were irradiated with defined doses of UVB and UVA and, subsequently, stored in Eppendorf tubes at –80°C until further use. UVA radiation was performed using a Dermalight Ultra1 type UVA1 24 kW phototherapy system (Dr Hönle, Munich, Germany; spectrum 340–440 nm, irradiance 80 mW per cm2 at a distance of 50 cm). UVB irradiation was carried out with a PL-S 9 W/12 (UV21) UVB light source (Philips, Aachen, Germany), with an emission peak at 313 nm. According to information supplied by the manufacturer, the UVC portion of this light source was less than 0.5%. The UVB irradiance was 0.33 mW per cm2 as measured using a Waldmann UV meter (Waldmann, Villingen-Schwenningen, Germany). All in vitro irradiation experiments were carried out on ice. Sebum-enriched Sebutapes® were extracted in 1 ml HPLC grade ethanol by vortexing for 1 min in 1.5 ml Eppendorf tubes. Then, tapes were removed and the remainder centrifuged at 2920×g and 4°C for 10 min 750 μl of each supernatant were transferred into an Eppendorf tube and directly subjected to HPLC analysis. The foreheads of healthy volunteers (n=4; all skin types III according to Fitzpatrick scale.) were divided into two equal areas. Prior to suberythematogenic exposures to UVA (20 J per cm2), and UVB (60 mJ per cm2), respectively, SSL of the left forehead were collected using a sterile cotton swab (Meditip® hygienic wooden applicators, Servoprax GmbH, Wesel, Germany) soaked in 100 μl ethanol. After UVA/UVB exposures, the same procedure was performed on the contralateral forehead area. Cotton applicator heads were transferred into 1.5 ml Eppendorf tubes and SSL extracted in 1 ml HPLC grade ethanol by vortexing for 1 min. Thereafter, cotton applicator heads were removed and samples centrifuged at 2920×g and 4°C for 10 min. Supernatant aliquots of 250 μl were immediately subjected to HPLC analysis. As this cotton swab method involves the use of ethanol, a highly volatile organic solvent, accurate weighing of very small amounts of SSL is not possible. Thus, for standardization purposes, squalene monohydroperoxide (SqmOOH) amounts detected in SSL extracts were related to the coeluted amounts of squalene. In the case of the experiments represented by Figure 6, this is valid because we have demonstrated that squalene levels are not significantly depleted by 20 J per cm2 UVA or 60 mJ per cm2 UVB (Figure 1).Figure 1Human sebum squalene is dose dependently depleted by UVA (B), but not by UVB (A). Sebum was collected from healthy volunteers using Sebutapes®, irradiated and immediately subjected to lipid extraction. Subsequently, squalene levels were measured by HPLC using UV detection as described. Shown are means±SEM, (A) n=7, (B) n=10. *p<0.05, **p<0.01; ***p<0.001 versus untreated control; n.s.=not significant.View Large Image Figure ViewerDownload Hi-res image Download (PPT) HPLC was performed using a Gynkotek HPLC system (Dionex-Softron GmbH, Germering, Germany). This system included an autosampler (Gina 50), pump (M480G), degasser, and UV/Vis detector (UVD 340, all from Gynkotek/Dionex-Softron) and a Luna 5μ C18 column (250×4.6 mm, Phenomenex®, Hösbach, Germany). The mobile phase consisted of HPLC grade ethanol and methanol (1:1, v/v). The flow rate was 1.2 ml per min. Control standards of squalene were prepared from commercially available pure squalene and was detected by in line UV detection at 210 nm. Peak integration and quantitation was performed by Gynkotek Software 5.6 (Dionex-Softron GmbH). For identification of unknown lipid peaks, fractions corresponding to these peaks (such as USLPP) were collected and subjected to further analysis. Atmospheric pressure chemical ionization mass spectrometry (APCI MS) spectra of squalene and USLPP were obtained on a PE Sciex API 165 single quadrupole mass spectrometer equipped with a Atmospheric Pressure Chemical Ionization (APCI) interface (Applied Biosystems, Langen, Germany). 1H nuclear magnetic resonance (NMR) spectra of USLPP were recorded on a Varian VXR 500 spectrometer equipped with a 2 mm probe, with approximately 100 μg of purified substance in a 40 μl NMR microcell. Spectra were measured in CD3OD with presaturation of the residual solvent peak. For the 1D spectrum, 1054 transients were accumulated, and mild line broadening function (0.25 Hz) applied prior to Fourier transformation. The adiabatic total correlation spectroscopy (Peti et al., 2000Peti W. Griesinger C. Bermel W. Adiabatic TOCSY for C,C and H,H J-transfer.J Biomol NMR. 2000; 18: 199-205Crossref PubMed Scopus (42) Google Scholar) spectrum was recorded using a mixing time of 0.065 s. 1H (500 MHz, CDCl3) δ: see Figure 2. APCI MS (positive ion mode) m/z 443 [M+H]+, 425 [(M+H) - 18]+, 409 [(M+H)-34]+. Squalene hydroperoxides were prepared by photo-oxidation. Squalene (20.0 g, 97%; Sigma-Aldrich) was dissolved in a mixture of methanol (40 ml) and benzene (160 ml), then rose Bengal sodium salt (150 mg) was added. The mixture was transferred to a rotating 1000 ml round flask, which was irradiated with a daylight bulb (250 W). The flask was kept at room temperature, and oxygen was bubbled through the reaction mixture at a flow rate of 15–20 l per h. After 8 h, the reaction was stopped and the solvent removed under reduced pressure. The residue was suspended in toluene (100 ml), and silica gel (40–63 μm; 15.0 g) was added for adsorption of rose Bengal. The suspension was left overnight, the solids filtered off and the solution evaporated under reduced pressure to provide an oily residue (20.7 g). A portion (19.7 g) of this residue was separated on a silica gel column (40–63 μm; 400×58 mm i.d.), eluted with benzene/ethyl acetate (93:7, v/v) with a flow rate of 7.5 ml per min. Fractions of 100 ml were collected, monitored by peroxide test and thin layer chromatography (TLC) analysis (details see below), and combined on the basis of a similar TLC pattern. The fraction containing the SqmOOH (elution volume 1200–2400 ml; 4.35 g) was further purified by preparative low pressure liquid chromatography on a Lobar reversed phase (RP)-18 column (40–63 μm, 440×37 mm i.d.; Merck, Darmstadt, Germany). The preparative chromatography system consisted of a K-1001 HPLC pump, a K-2501 UV/Vis detector (both Knauer, Berlin, Germany), a six-port injector, a LKB SuperFrac fraction collector and a LKB Rec-2 chart recorder (both Pharmacia, Freiburg, Germany). Aliquots were dissolved in 20 ml of a methanol/ethanol mixture (1:1, v/v) and separated with the same eluent. Flow rate was 6 ml per min, and detection was at 230 nm. Fractions of 12 ml were collected and analyzed by on a Hypersil ODS HPLC column (5 μm, 250×4 mm i.d.) with methanol/ethanol (1:1, v/v) at a flow rate of 1 ml per min and detection at 210 nm. The HPLC system consisted of a HP 1050 pump, HP 1040M Series II photodiode array detector, and a HP Chemstation (Agilent, Waldbronn, Germany). Fractions were combined on the basis of the HPLC profiles. Purification of 1.8 g of crude SqmOOH afforded a purified SqmOOH mixture (500 mg; RP HPLC purity>95%). 1H NMR (400 MHz, CDCl3) δ: 1.29 (CH3–OOH), 1.56 (CH3), 1.95–2.07 (CH2–CH=CH), 2.68 (HOO–CH–CH2–CH=CH), 4.28 (–CH–OOH), 5.00 (CH2=C), 5.05 (–CH=CH–), 5.5–5.6(–CH=CH–CH–OOH), 7.30 (C–OOH), 7.82 (C–OOH), 7.88 (C–OOH). APCI MS (positive ion mode) m/z 443 [M+H]+, 425 [(M+H)-18]+, 409 [(M+H)-34]+. Analysis was carried out on silica gel 60 F254 coated HPTLC plates (Merck). The sample concentrations were 1 mg per ml for squalene, purified squalene hydroperoxides from sebum, and for the synthetic squalene hydroperoxide mixture; sample concentrations for sebum samples were 10 mg per ml. Sample volumes of 6 μl (15 μl for squalene hydroperoxides from sebum) were applied on to the HPTLC plate with the aid of a AS 30 sample applicator (Desaga, Heidelberg, Germany). The plates were developed with benzene/ethyl acetate (93:7, v/v) over a distance of 9 cm. Visualization of compounds was performed using Godin's reagent (Godin, 1954Godin P. A new spray reagent for paper chromatography of polyols and cetoses.Nature. 1954; 174: 134Crossref Scopus (174) Google Scholar) followed by heating of the plate at 105°C for 2 min, and by spraying with N,N-DPDD reagent (Smith and Hill, 1972Smith L.L. Hill F.L. Detection of sterol hydroperoxides on thin-layer chromatoplates by means of the Wurster dyes.J Chromatogr. 1972; 66: 101-109Crossref PubMed Scopus (132) Google Scholar; Jork et al., 1993Jork H. Funk W. Fischer W. Wimmer H. Dünnschichtchromatographie: Reagenzien und Nachweismethoden. Vol.1b. VCH Verlagsgesellschaft, Weinheim1993Google Scholar). Statistical analysis was carried out by repeated measures paired one-way ANOVA (Graph Pad Instat®, Graph Pad Software, Inc., San Diego, CA) and a Bonferroni post-test. Human sebum squalene levels were not significantly depleted by UVB irradiation (Figure 4a), but were decreased by UVA in a concentration-dependent manner (Figure 1b). A significant reduction of sebum squalene levels to 71% of the initial squalene level was found after a single dose of 40 J per cm2 UVA (n=10, p< 0.05). Parallel to the decrease of squalene the generation of an initially USLPP was observed (Figure 2). To identify USLPP, samples of UV-irradiated sebum were analyzed by APCI MS. In APCI, analyte ions are generated at atmospheric pressure via chemical ionization in the gas phase. This ionization technique is suited for HPLC-MS analysis of lipophilic and poorly functionalized molecules such as squalene and other isoprenoids (Zhou and Hamburger, 1996Zhou S. Hamburger M. Application of liquid chromatography-atmospheric pressure ionization mass spectrometry in natural product analysis. Evaluation and optimization of electrospray and heated nebulizer interfaces.J Chromatogr A. 1996; 755: 189-204Crossref Scopus (83) Google Scholar). A typical chromatogram obtained with UV detection at 210 nm is shown in Figure 2, together with APCI mass spectra of SqmOOH (peak at 4.50 min) and squalene (peak at 9.50 min), recorded on-line. The mass spectrum of SqmOOH showed a weak quasi molecular ion at m/z 443, and diagnostic fragment ions at m/z 425 and m/z 409 resulting from the elimination of water and hydrogen peroxide from the parent ion. In contrast, the spectrum of the squalene peak showed a single protonated ion at m/z 411. USLPP was purified by repeated injections of sebum samples. The small amount of ≈ 100 μg of purified material available precluded extensive NMR experiments, and, in particular, the measurement of 13C spectra. The 1H signals in the 1D spectrum were assigned with the aid of adiabatic H,H total correlation spectroscopy experiment (Peti et al., 2000Peti W. Griesinger C. Bermel W. Adiabatic TOCSY for C,C and H,H J-transfer.J Biomol NMR. 2000; 18: 199-205Crossref PubMed Scopus (42) Google Scholar), comparison with reference spectra of squalene, and known chemical shift rules (Pretsch et al., 2000Pretsch E. Bühlmann P. Affolter C. Structure Determination of Organic Compounds: Tables of Spectral Data. 3rd. edn. Springer, Berlin2000Crossref Google Scholar). The assignments to the three different hemiterpene substructures occurring in the SqmOOH mixture are shown in Figure 3. In view of structural confirmation of USLPP and for further investigation of its biologic properties, access to larger amounts was required. SqmOOH was synthesized by photo-oxidation of squalene using a white light source and rose Bengal as a photosensitizer. Column chromatography of the crude squalene hydroperoxide reaction mixture on silica gel afforded a SqmOOH fraction, which was purified by RP chromatography to 97% purity. The analytical HPLC chromatogram of the synthetic material is shown in Figure 2(a). The retention time in RP HPLC was identical with that of an authentic sample of USLPP in irradiated sebum (Figure 2a). 1H NMR, APCI MS, and ESI MS data were comparable with those obtained for USLPP, which, hence, was identified as SqmOOH. A negative correlation between SqmOOH and squalene was found upon exposures of sebum to increasing doses of UVA (correlation coefficient r=–0.97; p < 0.005; Figure 4). Very low baseline levels of SqmOOH (0.765±0.23 nmol SqmOOH per mg sebum, Figure 5) were observed in samples of facial sebum collected by using Sebutapes®. Owing to a lack of significant squalene depletion (Figure 1a), and only very moderate formation of SqmOOH (Figure 5a) detected upon UVB exposures, no such correlation was obtained for UVB. SqmOOH levels increased slightly after exposure to UVB, however, more dramatically to UVA (Figure 5). Remarkably, formation of SqmOOH was found to be a highly sensitive marker of photo-oxidation particularly when compared with detection of squalene depletion: even very low suberythemogenic doses of UVA (5 J per cm2) increased SqmOOH levels significantly (UVA: 14.2±5.3 nmol SqmOOH per mg sebum; Figure 5b), whereas higher, erythemogenic doses of UVB only led to a slight increase of SqmOOH levels (UVB: 4.9±2.8 nmol SqmOOH per mg sebum; Figure 5a). Using a different method for extraction of SSL by cotton swabs soaked in ethanol, SqmOOH levels were determined in human SSL prior to and after in vivo irradiation. The low baseline levels (31±10 pmol SqmOOH per mmol squalene; Figure 6) before irradiation, as well as the highly increased levels upon a single exposure to 20 J per cm2 UVA (864±114 pmol SqmOOH per mmol squalene) confirmed findings from the in vitro experiments using Sebutapesâ. Preliminary tests with various solvent mixtures showed that best separations were achieved by combination of normal phase TLC on silica gel with benzene/ethyl acetate mixtures as eluents. Benzene/ethyl acetate (93:7, v/v) was finally chosen, as a suitable fingerprint of the entire spectrum of sebum lipids was obtained by single development. Detection of SqmOOH in UVA-irradiated sebum was straightforward without the need for tedious sample preparation prior to analysis by HPLC or TLC. The presence of SqmOOH in sebum was detected in as little as 60 μg of material. Lane 1 in Figure 7(a) shows a representative chromatogram of untreated sebum. The large spot at the top of the chromatogram (Rf 0.7) represents squalene. Other major spots at Rf 0.6 and 0.4 corresponded to sterol esters and fatty acids, the spot near the starting line (Rf 0.1) to polar lipids. Tentative identifications were based on chromatographic mobility relative to squalene and SqmOOH, and on the characteristic staining of fatty acids and isoprenoids with Godin's reagent. The UV-irradiated sebum (lane 2) showed a distinctly different pattern in the Rf range between 0.23 and 0.37, where several bands appeared. A purified sample of "USLPP" (lane 3) produced an identical pattern of at least five bands. The chromatogram of synthetic SqmOOH mixture (lane 4) had a fingerprint comparable with that of "USLPP". Given the much higher concentration applied here, minor spots from polar side products were detected between Rf 0 and 0.17). The staining reagent used in Figure 7(a), Godin's reagent, is a rather unspecific but all-purpose reagent for staining of isoprenoids and various lipids. Compounds containing a hydroperoxy moiety were selectively stained with a redox dye. Figure 7(b) shows the identical chromatogram stained with N,N-DPDD reagent (Smith and Hill, 1972Smith L.L. Hill F.L. Detection of sterol hydroperoxides on thin-layer chromatoplates by means of the Wurster dyes.J Chromatogr. 1972; 66: 101-109Crossref PubMed Scopus (132) Google Scholar). Peroxides oxidize the colorless aromatic diamine to a pink semiquinone diimine (Jork et al., 1993Jork H. Funk W. Fischer W. Wimmer H. Dünnschichtchromatographie: Reagenzien und Nachweismethoden. Vol.1b. VCH Verlagsgesellschaft, Weinheim1993Google Scholar). Lanes 2–4 show a pattern of stained bands at Rf 0.23–0.37, confirming that they all bear a hydroperoxy function. No N,N-DPDD-positive bands are observed