It has been postulated that reactive oxygen species gener-ated by UV irradiation contribute to photo-induced skindamage, such as lipid peroxidation, sunburn reaction, photo-toxicity, photoallergy, and photo-aging.1—8) However, themechanistic details remain unclear. Among various reactiveoxygen species, singlet oxygen, a highly reactive and toxicintermediate with specific reactivity, appears to play an im-portant role because it mediates photosensitized reac-tions.9,10) We previously constructed a very sensitive near-in-frared emission spectrometer with a germanium (Ge)-detec-tor as a reliable means of singlet oxygen detection.11—13) Weused this device to show that coproporphyrin from Propioni-bacterium acnes (P. acnes) generated singlet oxygen on thesurface of healthy skin under UV exposure by detecting theemission at 1268 nm due to singlet oxygen.13) Patients witherythropoietic protoporphyria, caused by accumulation ofprotoporphyrin in the skin, are known to suffer severe skindamage during sunlight exposure.14,15) Our data suggestedthat damage might also occur in healthy skin. In addition,UVA-induced singlet oxygen generation has been shown toinduce lipid peroxidation and gene expression.16—19) Minamiet al. reported that singlet oxygen was involved in UVA-in-duced oxidation of oleic acid and linoleic acid in mouse skin.However, singlet oxygen generated from coproporphyrin onthe human skin surface should immediately attack skin sur-face lipids, which are mainly derived from sebaceous glandlipids such as triacylglycerol, free fatty acid, wax, squaleneand cholesterol.16,20) Because squalene was hardly oxidized invitro under UV irradiation alone, a contribution of a factorproducing singlet oxygen to squalene peroxidation in theskin was suggested.21) Thus, there have been several investi-gations of singlet oxygen and endogenous photosensitizers inthe skin under physiological conditions.22—24) Lipid peroxideinduces further reactions and is associated with the pathologyof acne, atopic dermatitis, psoriasis and pigmentation.25,26)

However, the precise involvement of lipid peroxide in thesepathologies has not been well investigated.

Here, we first examined the reactivity of squalene with sin-glet oxygen directly by detection of singlet oxygen emission

at 1268 nm and clarified the contribution of coproporphyrinas a photosensitizer to squalene peroxide formation underUV exposure. Then we demonstrated that squalene peroxideparticipates in UVA-induced skin hyperpigmentation.

MATERIALS AND METHODS

Materials Squalene, cholesterol, oleic acid, linoleic acid,eosin yellowish and methylene blue were from Tokyo KaseiKogyo Co., Ltd. (Tokyo, Japan). Coproporphyrin III dihy-drochloride was from Frontier Scientic, Inc. (Logan, U.S.A.).Hematoporphyrin was from Sigma Chemical Co. (St. Louis,U.S.A.). Rose bengal was from Wako Chemical (Osaka,Japan). Kernechtrot stain solution was from Muto PureChemicals Ltd. (Tokyo, Japan). Other chemicals were com-mercial products of reagent grade.

Cell Culture Human epidermal melanocytes were cul-tured in the MEDIUM 254 supplemented with the humanmelanocyte growth supplement (HMGS) at 37 °C in 5%CO2/95% air condition. Human epidermal keratinocytes werecultured in the HuMedia-KG2 with 0.03 mM CaCl2, and hy-drocortisone was removed from media 72 h before treat-ments. Cells and media were from Kurabo Industries Ltd.(Osaka, Japan).

Animals Five-week-old female A-1 strain brownishguinea pigs were purchased from Tokyo Laboratory AnimalsScience Co., Ltd. (Tokyo, Japan) and housed under con-trolled temperature (22�1 °C) and humidity (50�10%) withaccess to standard food and water ad libitum. After an ac-climatization period of one week, the animals were subjectedto the following experiments.

Singlet Oxygen Emission Measurement The experi-mental setup consisted of an argon (Ar) laser (Inova 70-4;Coherent Inc., U.S.A.) and a near-infrared Ge-detector(model EI-S; Edinburgh Instruments Ltd., U.K.) cooled byliquid nitrogen and connected to the exit slit of a monochro-mator (model CT10; JASCO, Japan) with a blaze wavelengthat 1250 nm to minimize grating loss. The Ar laser output at514.5 nm was chopped at 800 Hz by an acousto-optic modu-

1504 Vol. 32, No. 9

Squalene as a Target Molecule in Skin Hyperpigmentation Caused bySinglet Oxygen

Akemi RYU,a Kumi ARAKANE,a Chiharu KOIDE,a Hiroyuki ARAI,b and Tetsuo NAGANO*,c

a Research Laboratories, KOSÉ Corporation; 1–18–4 Azusawa, Itabashi-ku, Tokyo 174–0051, Japan: b Department ofHealth Chemistry, Graduate School of Pharmaceutical Sciences, The University of Tokyo; and c Laboratory of Chemistryand Biology, Graduate School of Pharmaceutical Sciences, The University of Tokyo; 7–3–1 Hongo, Bunkyo-ku, Tokyo113–0033, Japan. Received March 27, 2009; accepted June 16, 2009; published online June 23, 2009

Based on our previous finding (Biochem. Biophys. Res. Commun., 223, 578—582, 1996) of singlet oxygengeneration from coproporphyrin excreted on the skin surface from Propionibacterium acnes, we hypothesizedthat singlet oxygen formed in this way under UV exposure would promote peroxidation of skin surface lipids. Wefound that squalene was oxidized efficiently by singlet oxygen derived from coproporphyrin under UV exposure,and that the rate constant of squalene peroxidation by singlet oxygen was ten-fold higher than that of other skinsurface lipids examined. The reaction was promoted more efficiently by UVA than by UVB. Furthermore, wefound that topical application of squalene peroxide induced skin hyperpigmentation through increasingprostaglandin E2 release from keratinocytes in guinea pigs. These results suggest that squalene peroxide forma-tion by singlet oxygen plays a key role in photo-induced skin damage.

Key words squalene peroxide; coproporphyrin; hyperpigmentation; photo-induced skin damage; singlet oxygen; UV

Biol. Pharm. Bull. 32(9) 1504—1509 (2009)

© 2009 Pharmaceutical Society of Japan∗ To whom correspondence should be addressed. e-mail: tlong@mol.f.u-tokyo.ac.jp

lator (A-160; Hoya, Japan) driven by a driver (110-DS;Hoya, Japan). The signal output from the Ge-detector wasfed to a model 124A lock-in amplifier via a model 116 pre-amplifier (both from E.G. & G Princeton Applied Research,U.S.A.) and synchronized with the internal reference signalof the lock-in amplifier. The signal output from the lock-inamplifier was fed to an XY recorder, and the emission spec-trum was recorded by scanning the grating with a motor.

Determination of Rate Constant for the Reaction ofSinglet Oxygen with Squalene The emission intensity ofsinglet oxygen generated in coproporphyrin (80 mM) ethanolsolution was monitored with or without squalene (0—80 mM)under excitation by Ar laser light at 514.5 nm with 200 mWoutput power. The rate constant (Kq) for the reaction of sin-glet oxygen with squalene was determined from Stern–Volmer plots, based on the equation I0/I�1�KqtCq, from theratio of the emission intensity with (I) or without (I0) squa-lene against the concentration of squalene (Cq) using the re-ported value for the life time of singlet oxygen (t) in ethanol(10—15�10�6 s).27,28)

UV Irradiation UV irradiation was performed with asolar simulator (XB-25T1W1, WACOM R&D Co., Tokyo,Japan) with UV cut filters that select the UVA region (320—400 nm) or UVB region (280—320 nm). UV output powerwas measured by an Eppley thermopile (The Eppley Lab.,Ind., U.S.A.). Each lipid (10 mM) ethanol solution was irradi-ated with UVA (6 mW/cm2) or UVB (0.15 mW/cm2) using asolar simulator in the presence of coproporphyrin (0—3 mM)or another photosensitizer (3 mM) on ice. From publisheddata, midday midsummer sunlight in Southern Europe isabout 5 mW/cm2.29)

Measurement of Lipid Peroxide Peroxide value (POV)was measured by the method described previously withpotassium iodide as a standard.30) Lipid peroxide was reactedwith iodide (I�), and tri-iodide (I3

�) was formed in equilib-rium I2�I�↔ I3

� with the excess I�. For the assay, chloro-form/acetic acid/ethanol�4 : 4 : 1 and a saturated potassiumiodide solution were prepared as solvents. These solventswere added to the lipid solutions exposed to UV light, andabsorbance of tri-iodide at 359 nm was measured after30 min.

Preperation of Squalene Peroxide Squalene peroxidewas prepared via the photosensitized reaction. Briefly, squa-lene (10 mM) and hematoporphyrin (1 mM) in ethanol solutionwere irradiated with UVA (10 J/cm2) using a solar simulator.The resulting solution contained 1.45 mM squalene peroxide/10 mM squalene as determined from POV measurement. Thesqualene peroxide mixture was used for further experimentswithout purification. Squalene (10 mM) and hematoporphyrin(1 mM) solution without irradiation were prepared as controls.

Topical Application of Squalene Peroxide The squa-lene peroxide mixture (10 mM) was applied on the left side ofthe shaved dorsal skin of four guinea pigs (24 nmol squaleneperoxide/cm2 per day, 5 d a week, for 3 weeks). Squalene(10 mM) and hematoporphyrin (1 mM) solution without irradi-ation were applied on the right side as controls. After 22 d,the animals were sacrificed and subjected to histological ob-servation. The experiments were conducted in accordancewith institutional guidelines for animal experiments at KOSÉCorporation.

Histological Observation of Skin Hyperpigmentation

Skin specimens were removed and fixed with phosphate-buffer (pH 7) containing 10% formaldehyde, and paraffin-embedded blocks were prepared. Skin sections of 5 mmthickness were prepared with a microtome and subjected toFontana–Masson staining as modified by Warkel.31) Nuclearcounterstaining was performed with Kernechtrot stain solu-tion. Image-Pro® Plus software (Media Cybernetics, Inc.,Maryland, U.S.A.) was used for image analysis.

Melanin Content Measurement Melanocytes weretreated with the squalene peroxide mixture (0.1—3 mM) orethanol, as a control, for 4 d and were assayed for melanincontent on the fifth day. After cell number was determinedwith a Coulter Counter (Beckman Coulter, Inc., Fullerton,U.S.A.), melanin was precipitated in 5% trichloroacetate anddissolved in 0.85 N KOH. The absorbance of the solution at400 nm was measured and melanin content was obtainedusing synthetic melanin as a standard.

Prostaglandin E2 (PGE2) Measurement Keratinocyteswere treated with the squalene peroxide mixture (12.5—50 mM) for 24 h and cells and media were collected sepa-rately. The solution of 50 mM squalene peroxide mixture con-tained 7.25 mM squalene peroxide. The squalene and hemato-porphyrin solution without irradiation were applied as con-trols. Cells were rinsed and homogenized with phosphatebuffer (pH 7.4), and the protein concentration of each cellhomogenate was determined using the BCA Protein AssayKit (Thermo Fisher Scientific Inc., Rockford, U.S.A.). Re-lease of PGE2 into media was determined using the pros-taglandin E2 EIA Kit (Cayman Chemical Company, AnnArbor, U.S.A.).

RESULTS

Determination of Reactivity of Squalene with SingletOxygen We previously showed that coproporphyrin, whichis a predominant porphyrin from P. acnes, on the skin surfacegenerated singlet oxygen under UV exposure, and that thepotential of coproporphyrin as a singlet oxygen generatorwas similar to that of other porphyrins and much higher thanthat of other photosensitizers, such as rose bengal and ri-boflavin.13) These suggested that human skin is always at riskof damage by singlet oxygen, and that squalene, one of themain components of sebaceous lipids, with six unsaturatedbonds, is readily oxidized by singlet oxygen on the skin sur-face.

First, the reactivity of squalene with singlet oxygen wasmeasured by detection of singlet oxygen emission in thepresence of various concentration of squalene. Figure 1shows the emission spectrum of singlet oxygen generatedfrom coproporphyrin and the decrease of emission intensityupon addition of squalene. The ratio of the emission intensitywith (I) or without (I0) squalene is plotted against the con-centration of squalene (Cq). From these Stern–Volmer plots,the value of the rate constant, Kq�2.6—3.9�106

M�1 s�1, was

obtained for squalene, based on the relationship I0/I�1�Kqt Cq. This value is in agreement with that measured byanother method.32) The reactivity of squalene with singletoxygen was more than ten-fold higher than reported valuesfor other skin surface lipids, such as oleic acid, linoleic acidand cholesterol (methyl oleate; Kq�1.3�105

M�1 s�1, methyl

linoleate; Kq�2.2�105M

�1 s�1, cholesterol; Kq�6.6�104

September 2009 1505

M�1 s�1).28)

Coproporphyrin Induced Squalene Peroxide Formationby UVA Irradiation Squalene peroxidation induced bysinglet oxygen was measured in the presence of various pho-tosensitizers, including coproporphyrin, under UVA expo-sure. From POV measurement, the amount of squalene per-oxide increased with increasing concentration of copropor-phyrin, and squalene was hardly oxidized in the absence ofcoproporphyrin (Fig. 2). On the other hand, UVB hardly pro-duced squalene peroxide in the presence of coproporphyrin.UVA was much more effective than UVB for producingsqualene peroxide at intensities proportional to those in solarlight (Fig. 3). Among the photosensitizers examined, copro-porphyrin and hematoporphyrin were the most efficient gen-erators of squalene peroxide, while rose bengal, eosin yel-lowish and methylene blue hardly enhanced squalene peroxi-dation (Table 1). Thus, coproporphyrin is a very effective en-dogenous photosensitizer for peroxidation of skin surfacelipid under UVA exposure. P. acnes-derived coproporphyrinexists in the sebaceous glands and excreted on the skin sur-

face with sebaceous lipid secretion. Then, peroxidation of se-baceous lipids by singlet oxygen was compared. Squalenewas much more easily oxidized by singlet oxygen than othersebaceous lipids such as cholesterol and unsaturated fattyacids (Table 2). These results are consistent with the findingthat the rate constant for the reaction of singlet oxygen withsqualene was much higher than those of the other lipids.These indicate that squalene is the first target of singlet oxy-gen from coproporphyrin.

Topical Application of Squalene Peroxide Induced Hy-perpigmentaion of Guinea Pig Skin As a model of photo-induced skin damage putatively mediated by squalene perox-ide, squalene peroxide prepared by photosensitized reactionwas applied to guinea pig skin. On the squalene peroxide-treated sides, sub-erythematic redness was observed 7—9 dafter application and this faded within 3—7 d (data notshown), followed by gradual appearance of pigmentation onday 18—22 in all four cases. Figure 4A shows a representa-tive example of the pigmented skin on day 20. Photomicro-graphs and their image analysis clearly showed increases in

1506 Vol. 32, No. 9

Fig. 1. Stern–Volmer Plots for Reaction of Singlet Oxygen with Squalene

Singlet oxygen was generated in coproporphyrin (80 mM) ethanol solution by Ar laserlight excitation at 514.5 nm with 200 mW output power. Singlet oxygen emission at1268 nm was monitored with (I) or without (I0) squalene. The rate constant of peroxi-dation of squalene (Kq) was determined from Stern–Volmer plots.

Fig. 2. Peroxidation of Squalene by Singlet Oxygen Produced by Copro-porphyrin

Squalene (10 mM) ethanol solution was irradiated with UVA (0—8.64 J/cm2) using asolar simulator in the presence of coproporphyrin (CP) (0—3 mM), and POV was meas-ured at the indicated time. Results are the mean�S.D. from three independent experi-ments.

Fig. 3. Peroxidation of Squalene on Exposure to UVA and UVB

Squalene (10 mM) ethanol solution was irradiated with UVA (0—8.64 J/cm2) or UVB(0—216 mJ/cm2) using a solar simulator in the presence of coproporphyrin (3 mM) andPOV was measured at the indicated time. Results are the mean�S.D. of three repli-cates.

Table 1. Comparison of Squalene Peroxidation Induced by Photosensitiz-ers

Photosensitizer DPOV (nmol lipid peroxide/mmol lipid)

Coproporphyrin 292.23�33.94Hematoporphyrin 242.71�18.47Rose bengal 15.47�1.96Eosin yellowish 4.71�3.78Methylene blue 2.39�0.08

Squalene (10 mM) ethanol solution was irradiated with UVA (8.64 J/cm2) using asolar simulator in the presence of photosensitizer (3 mM) and the increase of POV(DPOV) was measured. Results are the mean�S.D. of three replicates.

Table 2. Peroxidation of Various Lipids by Singlet Oxygen

Lipid DPOV (nmol lipid peroxide/mmol lipid)

Squalene 350.93�69.49Cholesterol 9.80�6.00Oleic acid 10.64�7.05Linoleic acid 9.48�4.52

Each lipid (10 mM) ethanol solution was irradiated with UVA (8.64 J/cm2) using asolar simulator in the presence of coproporphyrin (3 mM) and the increase of POV(DPOV) was measured. Results are the mean�S.D. from three independent experi-ments with three replicates.

the epidermal thickness and in the amount of melanin in thesqualene peroxide-treated skin (Figs. 4B—D). Squalene per-oxide appeared to activate keratinocyte proliferation andmelanin synthesis of melanocytes, resulting in skin hyperpig-mentation.

Effect of Squalene Peroxide on Melanin Synthesis ofCultured Melanocytes We examined the direct effect of squalene peroxide on melanin synthesis of culturedmelanocytes. Melanocytes were treated with squalene per-oxide mixture (0.1—3 mM) for 4 d. Squalene peroxide did not enhance but rather decreased melanin synthesis ofmelanocytes (Fig. 5). These indicated that the skin hyperpig-mentation caused by topical application of squalene peroxidewas not induced by its direct effect on melanocytes. It ishighly possible that squalene peroxide affects keratinocytesto release inflammatory factors inducing melanin synthesisof melanocytes.

Release of PGE2 from Keratinocytes Treated withSqualene Peroxide PGE2 is produced abundantly by ker-atinocytes in the skin on UV exposure and activatesmelanocyte dendricity and its melanin synthesis.33—36) Wethen measured PGE2 release from cultured keratinocytestreated by squalene peroxide. Keratinocytes were treated bysqualene peroxide mixture (12.5—50 mM) for 24 h. Squalene

peroxide dramatically increased PGE2 release from ker-atinocytes in a dose-dependent manner (Fig. 6). These indi-cated that squalene peroxide caused skin hyperpigmentation

September 2009 1507

Fig. 4. Effect of Topical Application of Squalene Peroxide Produced by Singlet Oxygen

Squalene (10 mM) and hematoporphyrin (1 mM) in ethanol were irradiated with UVA (10 J/cm2) using a solar simulator to afford a mixture containing squalene peroxide. Themixture was applied on the left side of the shaved dorsal skin of four guinea pigs (24 nmol/cm2 per day, 5 d a week, for 3 weeks). Squalene (10 mM) and hematoporphyrin (1 mM) so-lution without irradiation were applied on the right side as controls. After 22 d, the animals were sacrificed and subjected to histological observation. (A) A representative photo-graph of squalene peroxide-treated skin (left) and control skin (right). (B) Representative photomicrographs of squalene peroxide-treated or control skin. Paraffin-embedded sec-tions were processed for Fontana–Masson staining. Kernechtrot solution was used for nuclear counterstaining. Arrows show the thickness of the epidermis and arrowheads showmelanin. The scale bar indicates 200 mm. (C) Epidermal thickness was determined with Image-Pro® Plus software. Epidermal thickness was measured at 10 randomized spots perfield. Relative epidermal thickness of control skin was set to 1. Results are mean�S.D. from 10 fields. (D) Melanin content was determined as Fontana–Masson staining positivearea per field with Image-Pro® Plus software. Relative melanin content of control skin was set to 1. Results are mean�SD from 20 fields. The statistical significance was deter-mined by Student’s t-test (∗∗ p�0.01).

Fig. 5. Effect of Squalene Peroxide on Melanin Synthesis of CulturedMelanocytes

Human epidermal melanocytes were treated with squalene peroxide mixture (0.1—3 mM) for 4 d. Melanin was precipitated in 5% trichloroacetate and dissolved in 0.85 N

KOH. The absorbance of the solution at 400 nm was measured and melanin content wasobtained using synthetic melanin as a standard. Results are the mean�S.D. of threereplicates. The statistical significance from solvent control was determined by Student’st-test (∗ p�0.05, ∗∗ p�0.01).

through increasing PGE2 release from keratinocytes.

DISCUSSION

Many investigators have tried to unravel the involvementof reactive oxygen species in various diseases, includingphoto-induced skin damage. Although the reactivity and tox-icity of singlet oxygen are well known, there has been littleevidence of generation of singlet oxygen in vivo, except forerythropoietic protoporphyria, tetracycline phototoxicity andphotodynamic therapy of cancer.14,15,37—39) We constructed asensitive Ge-detector for detection of singlet oxygen emis-sion, and obtained evidence of singlet oxygen generation onthe surface of healthy skin at 1268 nm.13) P. acnes are normalinhabitants of human skin, and excrete porphyrins on theskin surface with sebaceous lipid secretion.40,41) Our previousreport showed that P. acnes-derived coproporphyrin producedsinglet oxygen under UVA excitation.13) This finding sug-gested that skin was always at risk of damage from singletoxygen generated by daily UV exposure, and that skin sur-face lipid was the first target of singlet oxygen.

In this study, we showed that squalene, a main componentof sebaceous lipid, had the highest reactivity with singletoxygen among the skin surface lipids examined, based on de-termination of the rate constant for the reaction of singletoxygen and squalene, using the most reliable detectionmethod for singlet oxygen. Further, coproporphyrin was apotent photosensitizer, inducing squalene peroxidation on theskin surface. In previous reports, high-intensity UV light orexogenous photosensitizers such as rose bengal were used tostudy squalene peroxide formation. We have established forthe first time that coproporphyrin is an efficient endogenousphotosensitizer that induces squalene peroxidation in theskin, and that UVA is much more effective than UVB forsqualene peroxide formation under physiological conditions.The reason may be that coproporphyrin has a larger absorp-tion in the UVA region than in the UVB region, in addition tothe fact that UVA has about ten times more energy than UVBin solar light. Our findings support the importance of protec-tion from not only UVB but also UVA to prevent skin dam-

age. Ekanayake Muduyanselage et al. investigated the gener-ation of squalene mono-hydroperoxide in sebaceous lipidsexposed to UVA/B under physiological condition, andshowed that UVA induced squalene peroxide at rates at leastone order of magnitude higher than did UVB.21) Since theskin surface is consistently exposed to UV in the presence ofcoproporphyrin, our observation of squalene peroxidationmediated by coproporphyrin is consistent with these in vivodata.

Although lipid peroxide is thought to be associated withthe pathogenesis of skin damage, such as acne, atopic der-matitis, psoriasis and pigmentation, the mechanisms involvedhave not been clarified in most cases.25,26) However, thepathological role of squalene peroxide in acne patients hasbeen investigated.42—46) It was reported that skin surfacelipids from acne patients were enriched in oxidized squalene,and that acne treatment decreased coproporphyrin in theskin.42,43) In this study, we demonstrated that singlet oxygenfrom coproporphyrin was most effective to squalene peroxideformation. The fact that P. acnes coproporphyrin is distrib-uted not only in acne patients, but also in healthy subjectsraises the possibility that squalene peroxide formation by sin-glet oxygen might mediate skin damage even in healthy sub-jects.

Indeed, we confirmed here that topical application of squa-lene peroxide produced by singlet oxygen induced hyperpig-mentation in guinea pig skin. However, squalene peroxidedid not enhance melanin synthesis of cultured melanocytes.A possible mechanism in the present case is as follows. Theincrease of melanin content induced by squalene peroxidecould be a result of the activation of melanin synthesis ofmelanocytes by inflammatory factors from keratinocytes.Squalene peroxide was reported to induce interleukin-6 se-cretion and lipoxygenase activation in HaCaT keratinocytesand was suggested to establish an inflammatory process viahyperkeratinization and sebogenesis in the pathogenesis ofacne.46) We found that squalene peroxide induced PGE2 re-lease from keratinocytes. Since PGE2 is known to activatemelanocyte dendricity and its melanin synthesis,33—36) squa-lene peroxide might cause skin hyperpigmentation throughstimulatting PGE2 release from keratinocytes. The concentra-tion of squalene peroxide in the experiment with ker-atinocytes was two orders of magnitude lower than that in thein vivo experiment, and much lower concentration of squa-lene peroxide decreased melanin synthesis of melanocytes.Considering the fact that melanocytes are located in the bot-tom layer of epidermal keratinocytes in the skin, it is possiblethat squalene peroxide on the skin surface has more signifi-cant effect on keratinocytes than melanocytes and, as a result,causes skin hyperpigmentation by stimulating PGE2 releasefrom keratinocytes. Our observation of hyperpigmentation inguinea pig skin supports this mechanism. Thus, we suggestedthat not only singlet oxygen itself, but also squalene perox-ide, has a key role in causing photo-induced skin damage.

Kohno et al. showed that the main oxidized squalene byUV irradiation on the skin surface was squalene mono-hydroperoxide, and that the amount of the squalene mono-hydroperoxide detected by HPLC corresponded quite well tothe POV, and determined that about 45 nmol squalene perox-ide from 1 mmol squalene was produced on the forehead aftersunlight exposure for 5 min.47,48) From the previous data that

1508 Vol. 32, No. 9

Fig. 6. Release of PGE2 from Keratinocytes Treated with Squalene Perox-ide

Human epidermal keratinocytes were treated with squalene peroxide mixture (12.5—50 mM) for 24 h. Squalene and hematoporphyrin solution without irradiation were ap-plied as controls. Release of PGE2 into media was determined using the prostaglandinE2 EIA Kit. Results are the mean�S.D. of three replicates. The statistical significancefrom solvent control was determined by Student’s t-test (∗ p�0.05, ∗∗ p�0.01).

sebaceous lipid on the forehead was about 150—300 mg/cm2

and that squalene was 12% of the sebaceous lipid,49) 44—88 nmol/cm2 of squalene existed on the forehead. Therefore,it was calculated that 2—4 nmol squalene peroxide/cm2/5 min was produced on the skin surface under daily sunlightexposure. Considering the sunlight exposure time, topical ap-plication of 24 nmol/cm2 of squalene peroxide in our experi-ments was not far away from those physiological conditions.

In conclusion, we have established that squalene is themost easily oxidized by singlet oxygen among the sebaceouslipids examined, and the rate constant of the peroxidationwas Kq�2.6—3.9�106

M�1 s�1. Further, coproporphyrin is a

potent endogenous photosensitizer for squalene peroxidationon the UVA-exposed skin surface. Moreover, topical applica-tion of squalene peroxide caused skin hyperpigmentation.These results indicate that squalene peroxide formation hasan important role in photo-induced skin damage.

REFERENCES

1) Dalle Carbonare M., Pathak M. A., J. Photochem. Photobiol. B, 14,105—124 (1992).

2) Niki E., Yoshida Y., Saito Y., Noguchi N., Biochem. Biophys. Res.Commun., 338, 668—676 (2005).

3) Sander C. S., Chang H., Salzmann S., Muller C. S., Ekanayake-Mudiyanselage S., Elsner P., Thiele J. J., J. Invest. Dermatol., 118,618—625 (2002).

4) Shin M. H., Rhie G. E., Kim Y. K., Park C. H., Cho K. H., Kim K. H.,Eun H. C., Chung J. H., J. Invest. Dermatol., 125, 221—229 (2005).

5) Yamamoto Y., J. Dermatol. Sci., 27 (Suppl. 1), S1—S4 (2001).6) Matsuo I., Yoshino K., Ohkido M., Curr. Probl. Dermatol., 11, 135—

143 (1983).7) Yasui H., Sakurai H., Exp. Dermatol., 12, 655—661 (2003).8) Takeshita K., Chi C., Hirata H., Ono M., Ozawa T., Free Radic. Biol.

Med., 40, 876—885 (2006).9) Dougherty T. J., Kaufman J. E., Goldfarb A., Weishaupt K. R., Boyle

D., Mittleman A., Cancer Res., 38, 2628—2635 (1978).10) Keene J. P., Kessel D., Land E. J., Redmond R. W., Truscott T. G., Pho-

tochem. Photobiol., 43, 117—120 (1986).11) Khan A. U., Kasha M., Proc. Natl. Acad. Sci. U.S.A., 76, 6047—6049

(1979).12) Kanofsky J. R., Basic Life Sci., 49, 211—218 (1988).13) Arakane K., Ryu A., Hayashi C., Masunaga T., Shinmoto K., Mashiko

S., Nagano T., Hirobe M., Biochem. Biophys. Res. Commun., 223,578—582 (1996).

14) Mathews-Roth M. M., Br. J. Dermatol., 134, 977 (1996).15) Mathews-Roth M. M., Methods Enzymol., 319, 479—484 (2000).16) Minami Y., Yokoyama K., Bando N., Kawai Y., Terao J., Free Radic.

Res., 42, 197—204 (2008).17) Klotz L. O., Holbrook N. J., Sies H., Curr. Probl. Dermatol., 29, 95—

113 (2001).18) Tyrrell R. M., Bioessays, 18, 139—148 (1996).19) Tyrrell R. M. “Oxidative Stress: Oxidants and Antioxidants,” ed. by

Sies H., Academic Press, San Diego, 1991, pp. 57—83.

20) Downing D. T., J. Dermatol. Sci., 62, 228—244 (1974).21) Ekanayake Mudiyanselage S., Hamburger M., Elsner P., Thiele J. J., J.

Invest. Dermatol., 120, 915—922 (2003).22) Sakurai H., Yasui H., Yamada Y., Nishimura H., Shigemoto M., Pho-

tochem. Photobiol. Sci., 4, 715—720 (2005).23) Baier J., Maisch T., Maier M., Landthaler M., Baumler W., J. Invest.

Dermatol., 127, 1498—1506 (2007).24) Hanson K. M., Simon J. D., Proc. Natl. Acad. Sci. U.S.A., 95, 10576—

10578 (1998).25) Hayakawa R., Ueda H., Nozaki T., Izawa Y., Yokotake J., Yazaki K.,

Azumi T., Okada Y., Kobayashi M., Usuda T., Ishida J., Kondo T.,Adachi A., Kawase A. Matsunaga K., Acta Vitaminol. Enzymol., 3,31—38 (1981).

26) Sugiura K., Iida Y., Ohgoshi H., Ueda H., Hirano K., Adachi T., Nip-pon Hifuka Gakkai Zasshi, 96, 165—170 (1986).

27) Nagano T., Arakane K., Ryu A., Masunaga T., Shinmoto K., MashikoS., Hirobe M., Chem. Pharm. Bull., 42, 2291—2294 (1994).

28) Wilkinson F., Brummer J. G., J Phys. Chem. Ref. Data, 10, 809—999(1981).

29) Diffey B., Robson J., J. Soc. Cosmet. Chem., 40, 127—133 (1989).30) Hicks M., Gebicki J. M., Anal. Biochem., 99, 249—253 (1979).31) Warkel R. L., Luna L. G., Helwig E. B., Am. J. Clin. Pathol., 73,

812—815 (1980).32) Kohno Y., Egawa Y., Itoh S., Nagaoka S., Takahashi M., Mukai K.,

Biochim. Biophys. Acta, 1256, 52—56 (1995).33) Imokawa G., Motegi I., J. Invest. Dermatol., 100, 47—54 (1993).34) Miller C. C., Hale P., Pentland A. P., J. Biol. Chem., 269, 3529—3533

(1994).35) Kabashima K., Nagamachi M., Honda T., Nishigori C., Miyachi Y.,

Tokura Y., Narumiya S., Lab. Invest., 87, 49—55 (2007).36) Scott G., Fricke A., Fender A., McClelland L., Jacobs S., Exp. Cell

Res., 313, 3840—3850 (2007).37) Weishaupt K. R., Gomer C. J., Dougherty T. J., Cancer Res., 36,

2326—2329 (1976).38) Li A. S., Chignell C. F., Hall R. D., Photochem. Photobiol., 46, 379—

382 (1987).39) Hasan T., Khan A. U., Proc. Natl. Acad. Sci. U.S.A., 83, 4604—4606

(1986).40) Cornelius C. E., Ludwig G. D. 3rd, J. Invest. Dermatol., 49, 368—370

(1967).41) Johnsson A., Kjeldstad B., Melo T. B., Arch. Dermatol. Res., 279,

190—193 (1987).42) Borelli C., Merk K., Schaller M., Jacob K., Vogeser M., Weindl G.,

Berger U., Plewig G., Acta Derm. Venereol., 86, 316—319 (2006).43) Saint-Leger D., Bague A., Lefebvre E., Cohen E., Chivot M., Br. J.

Dermatol., 114, 543—552 (1986).44) Motoyoshi K., Br. J. Dermatol., 109, 191—198 (1983).45) Chiba K., Yoshizawa K., Makino I., Kawakami K., Onoue M., J. Toxi-

col. Sci., 25, 77—83 (2000).46) Ottaviani M., Alestas T., Flori E., Mastrofrancesco A., Zouboulis C.

C., Picardo M., J. Invest. Dermatol., 126, 2430—2437 (2006).47) Kohno Y., Sakamoto O., Tomita K., Horii I., Miyazawa T., J. Oleo.

Sci., 40, 715—718 (1991).48) Kohno Y., Sakamoto O., Nakamura T., Miyazawa T., J. Oleo. Sci., 42,

44—49 (1993).49) Greene R., Downing D. T., Pochi P. E., Strauss J. S., J. Invest. Derma-

tol., 54, 240—247 (1970).

September 2009 1509