Article pubs.acs.org/crt
Oxidative Tryptophan Modification by Terpene- and SqualeneHydroperoxides and a Possible Link to Cross-Reactions in Diagnostic Tests Andreas Natsch,*,† Roger Emter,† Remo P. Badertscher,† Gerhard Brunner,† Thierry Granier,† Susanne Kern,† and Graham Ellis‡ †
Biosciences, Analytical Chemistry and Process Research Chemistry, Givaudan Schweiz AG, Ueberlandstrasse 138, CH-8600 Duebendorf, Switzerland ‡ RAPS Fragrance Toxicology, Givaudan International SA, 5 Chemin de la Parfumerie, CH-1214 Vernier, Switzerland S Supporting Information *
ABSTRACT: Hydroperoxides can act as specific haptens and oxidatively modify proteins. Terpene hydroperoxides trigger unusually high frequencies of positive skin reactions in human patients if tested at high concentrations. It is unknown whether this is due to specific hapten formation. Here, we show that both terpene hydroperoxides and the endogenous hydroperoxide formed from squalene can oxidatively modify tryptophan. Oxidative modifications of Trp were recently postulated to explain cross-sensitization between unrelated photosensitizers. Current observations may extend this hypothesis: Oxidative events triggered by endogenous hydroperoxides and hydroperoxides/oxidants derived from xenobiotics might lead to a sensitized state detected by patch tests with high concentrations of hydroperoxides.
■
INTRODUCTION Hydroperoxides 1 and 2 derived from the terpenes limonene and linalool (Scheme 1) are skin sensitizers in animals.1,2 A high prevalence of positive patch test reactions (5−7.2%) was found in dermatological clinics3−6 along with high frequency of doubtful/irritant reactions (6.4−10.5%), surpassing the prevalence of positive reactions to all organic skin sensitizers tested in baselines series.7 However, these frequencies were achieved only by testing at high concentrations.4 This finding is striking, as no data on significant exposure of the public to oxidized terpenes are available even though the parent compounds linalool and limonene have been found in a majority of consumer products for decades.8 Our recent analytical studies examining aged perfumes retrieved from consumers could, for the first time, identify trace levels of hydroperoxides in consumer products, but these levels were 3 to 4 orders of magnitude below the levels used in patch testing9,10 and 3 to 4 orders of magnitude below the inducing concentration in animal tests.2 Therefore, we concluded that these low levels do not represent a likely source of exposure sufficient to induce skin sensitization in a large fraction of dermatological patients. One possibility is that patch tests used to different terpene hydroperoxides do not reveal a hapten-specific sensitization state but, rather, a reactive state to other or multiple oxidizing agents. Thus, it is striking that a plot of the published frequency of positive and doubtful reactions to 1 and 2 reveals a high © 2015 American Chemical Society
correlation at the population level (Figure 1; meta-analysis of data from refs 5 and 6); however, this is only a tentative indication for a high level of concomitant reactions or crosssensitizations, as the data needs to be evaluated at the individual level. Individual-specific data have, so far, not been made publicly available.3,5,6 In the skin, the predominant unsaturated hydrocarbon is squalene, which is present at high concentrations and subject to UV-induced oxidation,11 leading to the formation of squalene hydroperoxide (3 is indicated here as one of the possible regioisomers). 3 is probably the major hydroperoxide naturally present in skin, but linoleic acid in skin lipids may also be prone to oxidation. An elegant recent study proposed a mechanism to explain why different photosensitizers lead to cross-sensitization in the absence of a chemically related haptenic structure.12 It was postulated that oxidative Trp modification leads to N-formylkynurenine, which may further react with Lys residues. Different photosensitizers can generate N-formyl-kynurenine by generating UV-induced singlet O2, oxidizing Trp. While such a process may also happen in the skin in the absence of xenobiotics, sufficiently immunogenic complexes may be formed by high levels of oxidative insult generated by Received: January 23, 2015 Published: May 5, 2015 1205
DOI: 10.1021/acs.chemrestox.5b00039 Chem. Res. Toxicol. 2015, 28, 1205−1208
Article
Chemical Research in Toxicology
Scheme 1. Structures of the Compounds Used in This Study and the Trp Oxidation Commonly Triggered by Multiple Oxidative Insults, with the Possible Link to the Proposed Haptenic Structure of Karlsson et al.12
and negative modes. For liquid chromatography separation, an XBridge C18 column with dimensions of 2.1 mm × 50 mm and particle size of 2.5 μm with a 2.1 mm × 10 mm precolumn of the same material (Waters) was used. The flow rate was 200 μL/min. Eluent A consists of water containing 0.1% formic acid, and eluent B consists of methanol containing 0.1% formic acid. A linear gradient was run from 90% eluent A (hold for 1 min) to 100% eluent B within 13 min (hold for 2 min), back to 90% eluent A within 2 min, followed by 2 min equilibration time. The injection volume of the sample was 1 μL. The resolution of the HR-MS spectra was set to 70 000. The mass accuracy was
