Introduction ORIGINAL PAPER

Transcription

1 Arch Dermatol Res (2003) 295 : DOI /s ORIGINAL PAPER Guillaume Bernard Elena Giménez-Arnau Suresh Chandra Rastogi Siri Heydorn Jeanne Duus Johansen Torkil Menné An Goossens Klaus Andersen Jean-Pierre Lepoittevin Contact allergy to oak moss: search for sensitizing molecules using combined bioassay-guided chemical fractionation, GC-MS, and structure-activity relationship analysis Received: 22 March 2003 / Revised: 13 June 2003 / Accepted: 22 July 2003 / Published online: 16 September 2003 Springer-Verlag 2003 G. Bernard E. Giménez-Arnau J.-P. Lepoittevin ( ) Laboratoire de Dermatochimie, Clinique Dermatologique, CHU, 1, Place de l Hôpital, Strasbourg, France Tel.: , Fax: , jplepoit@chimie.u-strasbg.fr S. C. Rastogi National Environmental Research Institute, Frederiksborgvej 399, 4000 Roskilde, Denmark S. Heydorn J. D. Johansen T. Menné Department of Dermatology, Gentofte Hospital, University of Copenhagen, 2900 Hellerup, Denmark A. Goossens Department of Dermatology, University Hospital, KU Leuven, Kapucijnenvoer 33, 3000 Leuven, Belgium K. Andersen Department of Dermatology, University Hospital, 5000 Odense C, Denmark Abstract In addition to pure synthetic fragrance materials several natural extracts are still in use in the perfume industry. Among them oak moss absolute, prepared from the lichen Evernia prunastri (L.) Arch., is considered a major contact sensitizer and is therefore included in the fragrance mix used for diagnosing perfume allergy. The process of preparing oak moss absolute has changed during recent years and, even though several potential sensitizers have been identified from former benzene extracts, its present constituents and their allergenic status are not clear. In the study reported here, we applied a method developed for the identification of contact allergens present in natural complex mixtures to oak moss absolute. The method is based on the combination of bioassay-guided chemical fractionation, gas chromatography-mass spectrometry analysis and structure-activity relationship studies. Our first results showed that atranol and chloroatranol, formed by transesterification and decarboxylation of the lichen depsides, atranorin and chloroatranorin, during the preparation of oak moss absolute, are strong elicitants in most patients sensitized to oak moss. Methyl-β-orcinol carboxylate, a depside degradation product and the most important monoaryl derivative of oak moss from an olfactory standpoint, was also found to elicit a reaction in most patients. Keywords Fragrance allergy Oak moss Atranol Chloroatranol Methyl-β-orcinol carboxylate Introduction Oak moss absolute is one of the eight ingredients of the fragrance mix included in the European standard patch test series as an indicator of fragrance contact allergy. Oak moss absolute is mainly derived from the lichen Evernia prunastri (L.) Arch. which grows primarily all over central and southern Europe, as well as in Morocco and Algeria. Because of its woody aroma and fixative properties, oak moss is used extensively in perfumery as a natural fragrance, particularly for masculine products such as after-shave lotions, cosmetics and fine perfumes [1]. Testing with the individual constituents of the fragrance mix has shown that oak moss absolute is one of the most frequent sensitizing ingredients [2, 3]. Oak moss absolute is traditionally prepared by extracting the harvested lichen with hydrocarbon solvents, and subsequently treating of the so-called oak moss concrete with a mixture of alcohols [1]. Several studies have been published on the chemical composition of oak moss absolute, of Evernia prunastri (L.) Arch. and of some other lichens [4, 5, 6, 7] as well as on the evaluation of the allergenic potential of the major components identified. So far, sensitivity to oak moss has been associated mainly with the presence of phenylbenzoates such as atranorin and evernic acid [8, 9, 10, 11]. Other compounds present in lichens, such as fumarprotocetraric acid, which have a phenylbenzoate functionality included in a cyclic system, have also been implicated in allergy [8, 11]. Usnic acid, although not a phenylbenzoate, has been shown to cause allergenic responses in some individuals [9, 10]. Also, compounds formed by transesterification of atranorin and chloroatranorin during the processing of oak moss, such

2 230 as ethyl chlorohematommate, have been shown to be allergenic in experimental models [12]. For many years benzene has been used to prepare the oak moss concrete. However, it is no longer used and has been replaced by more polar hydrocarbons. As a consequence, the composition of oak moss absolute has changed, and, even though several potential sensitizers have been identified from former benzene extracts, the present constituents of oak moss absolute and their allergenic status are not clear. Further, an increase in the number of eczema patients presenting with contact allergy to oak moss absolute has been reported over recent decades [13]. Therefore, there is a need to identify the main chemicals responsible for oak moss absolute contact allergy. In the course of our investigations on fragrance chemical allergy we have developed a new approach for the identification of fragrance sensitizers present in commercial perfumes and eaux de toilette [14, 15]. The method is based on the combination of bioassay-guided chemical fractionation, chemical analysis and structure-activity relationship (SAR) studies. Like commercial eaux de toilette and perfumes, a natural extract such as oak moss absolute contains several hundred different chemicals that are responsible for the complexity of the odour. We report here the first results obtained on the identification of contact allergens present in a natural complex mixture such as oak moss absolute. Materials and methods Chemicals Oak moss absolute samples were purchased from Biolandes Parfumerie (Grasse, France), Laboratoire Monique Rémy (Grasse, France), Mane & Fils (Le Bar sur Loup, France) and two from Prodarom (Syndicat National des Fabricants de Produits Aromatiques, Grasse, France). The oak moss absolute sample used for the fractionation study was the one from Biolandes Parfumerie. Dichloromethane (99.5%), acetone (99.8%) and methanol (99.9%) were purchased from Carlo Erba (Val de Reuil, France). Pentane (95%) was purchased from SDS (Peypin, France), and diethyl ether from Merck (Fontenay sous Bois, France). All solvents were used as delivered. Commercial atranorin, which is approximately a 1:1 mixture of atranorin and chloroatranorin, was purchased from Sigma- Aldrich (St Louis, Mo.). Orcinol (98%), β-orcinol (95%) and methyl-β-orcinol carboxylate (99%) were purchased from Acros Organics (Noisy le Grand, France), Sigma-Aldrich (St Quentin Fallavier, France) and Lancaster (Bischheim-Strasbourg, France), respectively. A mixture of atranol/chloroatranol (8:2) used for patch testing was prepared in the laboratory (unpublished results) from a sample of Yugoslavian oak moss absolute kindly provided by Chemotechnique Diagnostics (Malmö, Sweden). All other chemicals were obtained from Sigma-Aldrich (St Quentin Fallavier, France) and used without further purification. Chemical fractionation Gel permeation chromatography of oak moss absolute (1 g) was performed on a Sephadex LH-20 (Sigma-Aldrich, St Quentin Fallavier, France) column ( mm) with a mixture of dichloromethane/acetone (1:1) as eluent, giving fractions F1 F5. Fraction F4 was chemically refractionated by column chromatography on silica gel (Merck; Geduran 60, mm). Solvents for elution were first a mixture of pentane/diethyl ether 8:2 to give F4a and F4b then 7:3 to give F4c then a mixture of dichloromethane/methanol 9:1 to give F4d. All fractionations were monitored by thin-layer chromatography (TLC) carried out on glass plates precoated with silica gel 60 F-254 (Merck, 0.25 mm film thickness). The solvent for TLC development was a mixture of pentane/diethyl ether 7:3 and spots were visualized under a UV lamp at 254 nm. The plates were then sprayed with an anisaldehyde/sulphuric acid solution and charred on a hot-plate. For each fraction obtained the solvent was removed by evaporation under reduced pressure at room temperature. GC-MS instrumentation GC-MS analysis was carried out on a Perkin Elmer GC Auto System XL instrument coupled with a Perkin Elmer TurboMass mass spectrometer. The gas chromatograph was equipped with a splitless injector and a CP-Sil-5CB column (50 m 0.32 mm i.d µm film thickness; Chrompack, Middelburg, The Netherlands). Helium was used as the carrier gas (35 ml/min). The following temperature programme was used: 40 C (column temperature at injection), 40 C to 140 C at 30 C/min, 140 C to 280 C at 4 C/min hold at 280 C for 3 min. Mass spectra were obtained by electron impact ionization at 70 ev, ion source temperature 180 C. The mass spectra of the GC peaks were compared with the reference mass spectra of the chemical substances present in the mass spectrum library. The match parameters chosen for the identification of a compound were spectrum fit >900 counts (out of 1000 counts for absolute fit) and purity >800. Patients and patch testing All human studies were reviewed by the appropriate ethics committee and were therefore performed in accordance with the ethical standards laid down in the Declaration of Helsinki. All persons gave their informed consent prior to their inclusion in the study. Originally, a total of 40 patients with a previously diagnosed sensitivity to oak moss and negative to colophony were recruited in three European dermatological clinics (Gentofte, Odense and Leuven). They were randomly patch tested with five different oak moss absolute samples (1% pet.) and asked to participate in subsequent testing with fractions. For ethical reasons, all five samples were not patch tested in all patients. Several sets containing three of the five samples were established. Each patient was patch tested with one of these sets such that statistically the five samples were tested equally in the total of 40 patients. Initially, 30 of the 40 patients were volunteer subjects for the fractionation study. However, with time, four of them decided to quit the patch test procedure until the end of the chemical fractionations. In Gentofte and Odense patch tests were performed using Finn Chambers and Scanpore tape applied to the upper back, followed by occlusion for 2 days and reading on days 2, 3 and 7. In Leuven, subjects were patch tested using Van der Bend Chambers applied to the back of the patients with Micropore and fixed with Mefix, followed by occlusion for 2 days and reading on days 2 and 3 (or day 4). All the readings were performed using the ICDRG guidelines. Fractions F1 F5 and F4a F4d, as well as pure atranol and chloroatranol, were patch tested at 1% pet. equivalent to oak moss. This means that the patch test concentration of the fractions was calculated such that it corresponded to the amount of the fraction present in a sample of oak moss absolute 1% pet. The patch test concentration for other chemicals was 0.1% pet. Results In order to select a clinically representative sample, five commercial oak moss absolutes from different producers were patch tested randomly in the three clinical centres. No significant differences in the elicitation potential were

3 Table 1 Results of patch testing with fractions F1 F5 in a total of 30 patients. Fractions were patch tested at 1% pet. equivalent to oak moss. The patch test concentrations were calculated such that they corresponded to the amount of the fraction present in a sample of oak moss absolute 1% pet. (the percentage values in parentheses indicate the w/w percentage of each fraction in oak moss absolute). Readings were taken on day 3 (the numbers in parentheses indicate percentage of patients) Test material Number of positive and doubtful reactions ? F1 (21%) 4 (13) 13 (43) 4 (13) 3 (10) F2 (6%) 1 (3) 6 (20) 8 (27) 7 (23) F3 (5%) 1 (3) 12 (40) 5 (17) 5 (17) F4 (40%) 6 (20) 12 (40) 4 (13) 6 (20) F5 (28%) 2 (7) 10 (33) 3 (10) 6 (20) observed between the five samples and patients reacting positively reacted strongly to all the samples. Therefore, only one of the five oak moss absolutes was chosen for further chemical analysis. Gel permeation chromatography on Sephadex LH-20 (molecular size-based separation) gave five fractions, F1 F5, that were patch tested at 1% pet. equivalent to oak moss in 30 volunteers (the results are shown in Table 1). A relatively high number of patients gave a strong positive (++) response to one or several of the fractions (range 20 43%). Moreover, six patients gave a clear extreme positive (+++) reaction to F4 and four patients to F1, with intense erythema, infiltration and vesicles. Severe reactions of this type were much less frequently observed for F2, F3 and F5. Only one patient did not react to any of these five fractions. Further analysis focused on fraction F4, which represented 40% (w/w) of oak moss absolute. The chemical composition of F4 was investigated by GC-MS and the mass spectrum of each GC peak obtained was compared to the reference mass spectra of the chemical substances present in the MS library. In this way, 35 GC peaks were assigned to a specific compound. However, as 231 can be seen in Fig. 1, only some of these compounds were present in a significant amount. The chemical name and the molecular structure of the main constituents of F4 are shown in Table 2. These molecules were subjected to a SAR analysis for identification of substructures known to be associated with a sensitizing potential. Thus, all the major molecules either belonged to the resorcinol subgroup or were susceptible to metabolization into a resorcinol structure (compound 1) with in addition an aldehydic function in compounds 4 and 5. As compound 1 could be metabolized into compound 2 and as compounds 7 and 8 were structural analogues of compound 6, we selected compounds 2 to 6 as potentially sensitizing molecules to be patch tested. In parallel, silica gel column chromatography of F4 (polarity-based separation) gave four new subfractions, F4a F4d, that were patch tested at 1% pet. equivalent to oak moss in 26 of the initial volunteers. These patients were also patch tested with the main constituents of F4 (0.1% pet.), previously identified by GC-MS and classified as potential sensitizers by SAR analysis (the results are shown in Table 3). At the same time, the chemical composition of F4a F4d was analysed by GC-MS. Compounds 3 and 6 were both found in F4a and F4b. The mixture 4/5 was principally found in F4d, although it was also present in F4c and F4b to a much lower extent. Compound 2 was present in F4d. Fraction F4d was the most eliciting fraction with a total of 17 patients having a positive reaction, followed by F4b with 14 positive patients and F4c with 12 positive patients. Only four patients showed some positive reaction to F4a. The number of weak (+) and extreme (+++) reactions did not help distinguish between the eliciting potential of the fractions. However, 38% of the patients had a strong (++) reaction to F4d. In parallel, a total of 19 patients (73%) had a positive reaction to the mixture chloroatranol/atranol 4/5 (2:8), and among these 6 developed a very strong reaction with intense erythema, infiltration and coalescing vesicles. These results explained the strong elic- Fig. 1 GC chromatogram of fraction F4

4 232 Table 2 Main constituents of fraction F4 (CAS Chemical Abstracts Service) Table 3 Results of patch testing with fractions F4a F4d and the main constituents of F4 in a total of 26 patients. Fractions were patch tested at 1% pet. equivalent to oak moss (the percentage values in parentheses indicate the w/w percentage of each fraction in oak moss absolute). Pure chemicals were patch tested at 0.1% pet. Readings were taken on day 3 (the numbers in parentheses indicate percentage of patients) Test material Number of positive and doubtful reactions ? F4a (20%) 1 (4) 1 (4) 2 (8) 2 (8) F4b (14%) 2 (8) 5 (19) 7 (27) 8 (31) F4c (3%) 1 (4) 4 (15) 7 (27) 6 (23) F4d (5%) 1 (4) 10 (38) 6 (23) 7 (27) 2 a 0 (0) 0 (0) 0 (0) 2 (8) 3 a 0 (0) 2 (8) 1 (4) 3 (12) 4/5 b 6 (23) 7 (27) 6 (23) 3 (12) 6 a 0 (0) 3 (12) 3 (12) 4 (15) a Commercially available b Chloroatranol 4 and atranol 5 are not commercially available. A mixture of 4/5 (2:8) previously obtained from a sample of Yugoslavian oak moss absolute was used iting potential of F4d and F4c, compound 2 being negative at patch testing. On the another hand, six patients (23%) were positive to compound 6 with three of them having a ++ response, and three (12%) to compound 3 with two of them having a ++ response. This explained the few positive reactions observed to F4a. Because fraction F4b was a mixture containing compounds 4/5, 3 and 6, it was not surprising that a high number of positive reactions were observed. One interesting case was observed in Leuven: a 59-year-old woman with strong positive responses to F4a, F4b and compound 6, was completely negative to the other fractions and to the mixture chloroatranol/atranol 4/5. Based on these results, the mixture chloroatranol/atranol 4/5 (2:8) was a strong elicitant of contact allergy to oak moss in already sensitized patients. Chloroatranol 4 and atranol 5 are not commercially available, so, in order to evaluate their eliciting potential separately, they were synthesized according to classical organic chemistry methods (Fig. 2). Afterwards, compounds 4 and 5 were patch tested at 1% pet. equivalent to oak moss in 15 of the 26 remaining patients. A first analysis of the results (Table 4) indicated that both compounds had a strong eliciting potential. A total of 10 (67%) and 11 (73%) patients showed a positive reaction to chloroatranol 4 and atranol 5, respectively. No significant differences were observed between the number of weak (+), strong (++) and extreme (+++) reactions. However, the percentage (w/w) of chloroatranol 4 in the oak moss sample was lower (0.9%)

5 Fig. 2 Synthesis of chloroatranol 4 and atranol 5 Table 4 Results of patch testing with chloroatranol 4 and atranol 5 in a total of 15 patients. The patch test concentration was 1% pet. equivalent to oak moss (the percentage values in parentheses indicate the estimated w/w percentage in oak moss absolute). Readings were taken on day 3 (the numbers in parentheses indicate percentage of patients) Test material than the concentration of atranol 5 (2.1%), which means that the patch test concentration, 1% pet. equivalent to oak moss, was lower for chloroatranol (90 ppm) than for atranol (210 ppm). Chloroatranol could therefore be considered a stronger elicitant than atranol. Discussion Number of positive and doubtful reactions ? 4 (0.9%) 1 (7) 2 (13) 7 (47) 3 (20) 5 (2.1%) 1 (7) 3 (20) 7 (47) 3 (20) For many years, oak moss absolute has been used mixed with tree moss. The presence in tree moss of resin acids, which are the major constituents of colophony, a wellknown skin sensitizer, has previously been reported [16]. From July 2001 it is stated in the International Fragrance 233 Association (IFRA) Code of Practice that oak moss extracts used in perfume compounds must not contain added tree moss. However, it is also indicated that traces of resin acids are unavoidable in current commercial qualities of oak moss due to contamination problems during the lichen collection. Therefore, only patients with a previous known sensitivity to oak moss but negative to colophony were recruited in order to avoid any source of misdiagnosis. At the beginning of the study, five commercial samples of oak moss absolute from different producers were selected for random patch testing in the three clinical centres, but no significant differences in the elicitation potential were observed between the five samples. The patients reacting positively reacted strongly to all the samples, and therefore only one of the five oak moss absolutes was chosen for further analysis. After a first round of fractionation and patch testing, we focused on fraction F4 that seemed to be the strongest elicitor in our patients. Chemical analysis identified eight chemicals derived from a resorcinol structure. Moreover, two of these chemicals, namely compounds 4 and 5, contained an additional aldehydic function. A SAR analysis indicated that they all should be considered potential sensitizers, either through an oxidation of the resorcinol system and/or by direct reaction of the aldehydic functions. In order to avoid too many patch tests, we selected the most representative molecules covering the different molecular structures. Thus, compound 1, a methylated analogue of compound 2, and compounds 7 and 8, ester analogues of compound 6, were not tested. Results obtained after a single chemical fractionation of the oak moss absolute sample, patch testing, and analytical and SAR studies, led to the identification of chloroatranol 4 and atranol 5 as major eliciting chemicals, and of methyl-β-orcinol carboxylate 6 and β-orcinol 3 as minor ones. Further patch testing with pure synthetic chloroatranol 4 and atranol 5 confirmed their strong eliciting potential with some patients reacting +++/++ at 90 and 210 ppm, respectively. In order to confirm this identification based on fractionation, and analytical and SAR analysis, fraction 4 was subfractionated into F4a F4d, and these subfractions were chemically analysed and patch tested. The results obtained were in full agreement with the identified molecules as the positive subfractions were the ones containing compounds 3, 4, 5 and 6. Several rounds of testing were therefore carried out. With so many exposures per patient, sensitization booster effects could be expected. However, there was absolutely no tendency for stronger reactions with further testing. On another hand, and in order to really avoid multiple exposure, reproducibility patch test assays were not performed. However, the fact that all patients who tested positive at the beginning continued to react to at least one of the subfractions and/or to at least one of the test pure substances is an argument for reproducibility. Freshly harvested oak moss has substantially no scent. The moss contains various types of depsides, which are non-volatile, odourless, polyfunctional diaryl derivatives [17]. The most important of these depsides are lecanoric acid, evernic acid, divaricatic acid, barbatic acid, atranorin,

6 234 Fig. 3 Depsides in oak moss chloroatranorin and thamnolic acid (Fig. 3). The characteristic oak moss fragrance is only developed after cleavage of the depsides during treatment of the oak moss concrete with alcohols to give volatile, scented, monoaryl derivatives [18]. For example, the transesterification and decarboxylation of atranorin and chloroatranorin during the ethanolic treatment is shown in Fig. 4. The monoaryl derivatives 4/6 are, therefore, degradation products of atranorin and chloroatranorin formed during oak moss processing. Resorcinol derivatives such as methyl-β-orcinol carboxylate 6 are mainly responsible for the characteristic earthy-moss-like odour of the oak moss products [19], and can also be derived from the decomposition of other depsides such as barbatic acid. Compound 3 could be formed by decarboxylation of compound 6. However, compound 3 was observed only with the GC-MS analysis and was not observed with other structural elucidation techniques such as nuclear magnetic resonance (data not shown). This suggested that compound 3 could be formed from compound 6 as a result of the high temperature of the GC. This was confirmed by reference GC-MS experiments using commercial atranorin, which showed approximately a 1:1 mixture of atranorin and chloroatranorin. GC peaks were indeed assigned for chloroatranol 4, atranol 5, methyl-βorcinol carboxylate 6 and β-orcinol 3, confirming the high thermal lability of atranorin and chloroatranorin. For a chemical to be a contact sensitizer, it must, after penetration into the epidermis, bind to skin proteins so Fig. 4 Degradation products of atranorin and chloroatranorin formed during oak moss processing that an antigen is formed [20]. From a chemical point of view, compounds 3 6 are 1,3-benzenediol derivatives that can be considered as prohaptens requiring a biotransformation to become sensitizing [21]. Aromatic hydroxyl functions may be oxidized enzymatically to form quinones, that are known sensitizers. In addition to the phenolic feature, chloroatranol 4 and atranol 5 also contain an aldehydic function. Aldehydes are highly reactive molecules, and many are strong sensitizers. However, it has been shown that aldehydes directly conjugated to an aromatic ring are weak sensitizers or nonsensitizers compared to those that contain a nonconjugated aromatic ring [21]. The enzymatic oxidation mechanism mentioned above could therefore act as an alternative for compounds 4 and 5. However, the precise structure of the final hapten derived from these prohaptens is not known. We describe here a method for the accurate identification of allergens in oak moss absolute. A combination of chemical fractionation, patch testing and analytical and SAR methods allowed the identification of chloroatranol 4 and atranol 5, formed by transesterification and decarboxylation of atranorin and chloroatranorin during the oak moss absolute derivatization, as offending chemicals. Other elicitants found, although to a lower extent, were the dep-

7 side degradation products methyl-β-orcinol carboxylate 6 and β-orcinol 3, the most important monoaryl derivatives of oak moss from an olfactory standpoint. For many years, sensitivity to oak moss was mainly correlated with a sensitivity to atranorin, evernic acid and fumarprotocetraric acid [8, 9, 10, 11]. Later, compounds formed by transesterification of atranorin and chloroatranorin during oak moss processing, such as ethyl chlorohematommate, were shown to be allergenic [12]. Hiserodt et al. [7] have recently reported the total chemical decomposition of atranorin and chloroatranorin. The results presented here indicate that monoaryl compounds derived from the decomposition of depsides are allergenic. Depsides still remaining in the absolute after treatment of the lichen with solvents may have been present in the first fractions eluted during gel permeation chromatography of the oak moss sample (molecular size-based separation), which were also positive at patch testing. Bioassayguided chemical subfractionation of these fractions, and analytical and SAR studies are currently underway. Acknowledgements The assistance of Mireille Huel for the synthesis of atranol and chloroatranol is gratefully acknowledged. This work was financially supported by the Fifth Framework Programme of the European Commission, under the Quality of Life and Management of Living Resources thematic programme, key action Environment and Health (Contract QLK4-CT : Fragrance chemical allergy: a major environmental and consumer health problem in Europe ). Research participants: Jean-Pierre Lepoittevin, Elena Giménez-Arnau and Guillaume Bernard, University Louis Pasteur, Strasbourg, France; Ann-Therese Karlberg, Mihaly Matura, Anna Börje and Maria Sköld, National Institute for Working Life, Stockholm, Sweden; David Basketter and Grace Patlewicz, SEAC, Unilever Research, Bedford, UK; Torkil Menné, Jeanne Duus Johansen and Siri Heydorn, Department of Dermatology, Gentofte Hospital, University of Copenhagen, Denmark; Peter J. Frosch, Department of Dermatology and University of Witten/Herdecke, Dortmund, Germany; Ian White, Department of Contact Dermatitis and Occupational Dermatology, St. John s Institute of Dermatology, St. Thomas Hospital, London, UK; Suresh Rastogi, National Environmental Research Institute, Roskilde, Denmark; Klaus E. Andersen, Department of Dermatology, Odense University Hospital, Denmark; Magnus Bruze and Cecilia Svedman, Department of Occupational Dermatology, University Hospital Malmö, Sweden; and An Goossens, Department of Dermatology, University Hospital KU Leuven, Belgium. References 1. Actander S (1960) Oak moss. In: Perfume and flavor materials of natural origin. Elisabeth, NJ, pp Johansen JD, Menné T (1995) The fragrance mix and its constituents: a 14-year material. Contact Dermatitis 32: Johansen JD, Heydorn S, Menné T (2002) Oak moss extracts in the diagnosis of fragrance contact allergy. Contact Dermatitis 46: Gavin J, Tabacchi R (1975) Isolement et identification de composés phénoliques et monoterpéniques de la mousse de chêne (Evernia prunastri (L.) Ach.). Helv Chim Acta 58: Gavin J, Nicollier G, Tabacchi R (1978) Composants volatils de la mousse de chêne (Evernia prunastri (L.) Ach.). Helv Chim Acta 61: Schulz H, Albroscheit G (1989) Characterization of oak moss products used in perfumery by high-performance liquid chromatography. J Chromatogr 466: Hiserodt RD, Swijter DFH, Mussinan CJ (2000) Identification of atranorin and related potential allergens in oak moss absolute by high-performance liquid chromatography-tandem mass spectrometry using negative ion atmospheric pressure chemical ionization. J Chromatogr A 888: Dahlquist I, Fregert S (1980) Contact allergy to atranorin in lichens and perfumes. Contact Dermatitis 6: Thune P, Solberg Y, McFadden N, Staerfelt F, Sandberg M (1982) Perfume allergy due to oak moss and other lichens. Contact Dermatitis 8: Sandberg M, Thune P (1984) The sensitizing capacity of atranorin. Contact Dermatitis 11: Gonçalo S, Cabral F, Gonçalo M (1988) Contact sensitivity to oak moss. Contact Dermatitis 19: Terajima Y, Ichikawa H, Tokuda K, Nakamura S (1988) Quantitative analysis of oak moss oil. In: Lawrence BM, Mookerjee BD, Willis BJ (eds) Flavors and fragrances: a world perspective. Elsevier Science, Amsterdam, pp Buckley DA, Wakelin SH, Seed PT, Holloway D, Rycroft RJG, White IR, McFadden JP (2000) The frequency of fragrance allergy in a patch test population over 17 years. Br J Dermatol 142: Mutterer V, Giménez Arnau E, Lepoittevin J-P, Johansen JD, Frosch PJ, Menné T, Andersen KE, Bruze M, Rastogi SC, White IR (1999) Identification of coumarin as the sensitizer in a patient sensitive to her own perfume but negative to the fragrance mix. Contact Dermatitis 40: Giménez Arnau E, Andersen KE, Bruze M, Frosch PJ, Johansen JD, Menné T, Rastogi SC, White IR, Lepoittevin J-P (2000) Identification of Lilial as a fragrance sensitizer in a perfume by bioassay-guided chemical fractionation and structureactivity relationships. Contact Dermatitis 43: Lepoittevin J-P, Meschkat E, Huygens S, Goosens A (2000) Presence of resin acids in Oakmoss patch test material: a source of misdiagnosis? J Invest Dermatol 115: Ter Heide R, Provatoroff N, Traas PC, De Valois PJ, Wobben HJ, Timmer R (1975) Qualitative analysis of the odoriferous fraction of oakmoss (Evernia prunastri (L.) Arch.). J Agric Food Chem 23: Boelens MH (1993) Formation of volatile compounds from oakmoss. Perfumer Flavorist 18: Bauer K, Garbe D, Surburg H (1997) Common fragrance and flavor materials: preparation, properties and uses, 3rd edn. Wiley-VCH Verlag, Weinheim 20. Rustemeyer T, Van Hoogstrate IMW, Von Blomberg BME, Scheper RJ (2001) Mechanisms in allergic contact dermatitis. In: Rycroft RJG, Menné T, Frosch PJ, Lepoittevin J-P (eds) Textbook of contact dermatitis, 3rd edn. Springer, Berlin Heidelberg, pp Smith CK, Hotchkiss SAM (2001) Xenobiotics as skin sensitizers: metabolic activation and detoxication, and protein-binding mechanisms. In: Smith CK, Hotchkiss SAM (eds) Allergic contact dermatitis: chemical and metabolic mechanisms. Taylor & Francis, London, pp