Chem. Res. Toxicol. 2009, 22, 1541–1547
1541
Chemical Mechanisms for Skin Sensitization by Aromatic Compounds with Hydroxy and Amino Groups Aynur O. Aptula,† Steven J. Enoch,‡ and David W. Roberts*,‡ Safety and EnVironmental Assurance Centre (SEAC), UnileVer, Colworth, Sharnbrook, Bedford, MK44 1LQ, England, and School of Pharmacy and Chemistry, LiVerpool John Moores UniVersity, Byrom Street, LiVerpool, L3 3AF, United Kingdom ReceiVed January 29, 2009
It is well-known that aromatic diamino-, dihydroxy-, and amino-hydroxy compounds, with NH2 and OH groups in ortho- or para-positions relative to each other, are strong skin sensitizers. In this paper, we analyze published potency and cross-reactivity data, whereby animals sensitized to one of these compounds are challenged with other compounds. The data are consistent with two parallel chemical reaction mechanisms: oxidation to electrophilic (protein reactive) quinones or quinone imines or formation of protein-reactive free radicals such as the Wu¨rster salt, which can be formed by para-phenylene diamine. Compounds with NH2 and OH groups meta to each other have also been found to be skin sensitizers, in some cases quite strong sensitizers. For these compounds, direct formation of quinones or quinone imines is not possible, and free radicals of the Wu¨rster salt type are not favored. Here, we present a molecular mechanism to rationalize the sensitization potential of such compounds and, using the results of quantum mechanics calculations, show how this mechanism can explain observed structure-potency trends. Introduction An animal testing ban on chemicals to be used in cosmetics came into effect in the European Union (EU) in March, 2009, linked to an EU marketing ban on products containing any ingredients that have been subsequently tested in animals for acute toxicity, skin irritation, eye irritation, and mutagenicity (1). For various repeat-dose toxicity tests, including skin sensitization, the EU marketing ban is due to come into effect in March, 2013 (1). This means that nonanimal methods for the prediction of skin sensitization potency need to be developed and implemented. Much progress has been made in mechanismbased methods for the prediction of skin sensitization based on reaction chemistry (2-10). The requirement for a chemical to covalently modify skin protein(s) has long been established as a key event in the induction of skin sensitization (11). The skin sensitization properties of most, if not all, known contact allergens can be rationalized in terms of electrophilicity or pro-electrophilicity of the allergens, enabling them to bind to proteins. More details on the determinants of skin sensitization potential can be found in ref 4. Multisubstituted aromatic compounds are among the most difficult to predict, exhibiting a range of possible chemical mechanisms via several potential activation pathways. Aromatic compounds of the general structure C6H4XY, where X and Y are hydroxy or amino groups, fall within this category and are the subject of the present paper. The strong skin sensitization potential of para-phenylenediamine (PPD) and hydroquinone (HQ) has been well-recognized for many years (12), but nevertheless, the mechanistic basis continues to be an active area of research (13, 14). HQ is usually * To whom correspondence should be addressed. Tel: +44/1512312422. Fax: +44/1512312170. E-mail:
[email protected]. † Unilever. ‡ Liverpool John Moores University.
considered to be a pro-electrophilic sensitizer (prohapten), being oxidized either abiotically or enzymatically to benzoquinone (BQ), which is chemically reactive as a Michael acceptor electrophile. BQ itself is recognized as an extreme sensitizer, and guinea pig cross-challenges (see later) support the view that HQ sensitizes via BQ (12). One of several possible chemical mechanisms for sensitization by PPD is similar to that described above for HQ: oxidation of PPD to a di-imine (DI), the nitrogen analogue of BQ, which can either act as a Michael acceptor electrophile or undergo hydrolysis to BQ. Recently, Aeby et al. (13) analyzed the role of oxidation and N-acetylation as major transformation steps, leading to activation and deactivation, respectively, in the sensitization mechanism of PPD. For ortho analogues of PPD and HQ, a similar chemistry is possible and can be invoked to explain the skin sensitization observed with such compounds. There is, however, evidence that the above mechanistic model may be oversimplistic, as was recognized by Basketter and Goodwin (12). First, it would predict that PPD, HQ, and BQ should be highly cross-reactive with each other, since all are proposed to act by the same “ultimate hapten”, BQ. Second, it would predict that aromatic compounds with amino or hydroxy groups meta to one another should not sensitize, since metaquinones or meta-quinone imines cannot exist except in the form of very short-lived diradicals. Both of these predictions are incorrect, as discussed in the next section. The aim of the present paper is to consider the mechanistic chemistry of aromatic diamino-, dihydroxy-, and amino-hydroxy compounds in more detail and, in particular, to consider the role of free radical chemistry in rationalizing the observed structure-activity trends for sensitization potential and crossreactivity.
Sensitization and Cross-Challenge Data In the earlier literature, cross-challenge studies involving dihydroxy-, diamino-, and amino-hydroxy-benzenes have been
10.1021/tx9000336 CCC: $40.75 2009 American Chemical Society Published on Web 08/13/2009
1542
Chem. Res. Toxicol., Vol. 22, No. 9, 2009
Aptula et al.
Table 1. TES Values for Challenges and Cross-Challenges in the GPMT challenge induction
PPD
HQ
BQ
PPD HQ BQ
90 0 15
7 34 71
80
Table 2. TES Values for Challenges and Cross-Challenges in the MSIAT column A
B
C
D
E
F
challenge row
induction
PPD
HQ
BQ
OAP
MAP
PAP
1 2 3 4 5 6
PPD HQ BQ OAP MAP PAP
57 0 8 11 27 40
3 60 40
33 43 57 13 16 36
20 0 8 23 13 21
25 0 0 0 39 34
20 0 0 6 26
reported using three guinea pig assays: the GPMT (guinea pig maximization test), the MSIAT (modified single injection adjuvant test), and the CCET (cumulative contact enhancement test) (12). The results are summarized in Tables 1-3. In cross-challenge studies, animals that have been sensitized by one compound are challenged with other compounds to see if the latter can elicit a sensitization response. For example, consider the MSIAT results for sensitization by meta-aminophenol (MAP), summarized in Table 2. Induction of sensitization was by intradermal injection of an MAP solution to 10 guinea pigs previously injected with Freund’s complete adjuvant followed by resting for 2 weeks. The animals were then challenged by application of a solution of MAP, occluded under a patch, to a small area of shaved skin, which was then examined after 24 and 48 h to assess the response by scoring the resulting erythema as 0 (no sensitization response), 0.5 (uncertain), 1, 2, or 3 (all indicating a definite sensitization response). Nine of the 10 animals scored 1 or higher, and the other animal scored zero. The average erythema score, over the nine positive animals, was 1.3. Because the maximum possible score for 10 animals is 30, the % total erythema score (TES) is 100 × 9 × 1.3/30 ) 39. At weekly intervals after this, the animals were challenged on alternate flanks, using the same patch procedure and scoring system, with solutions of PPD, BQ, and orthoaminophenol (OAP) [it is not stated in the original reference (12) how many of these challenges were done at the same time, using different areas of skin]. PPD produced a response in eight animals, with an average erythema score of 1.0, so the TES for this challenge is 100 × 8 × 1/30 ) 27. The TES values shown in Table 2 for the MAP-sensitized animals cross-challenged with BQ and OAP were determined similarly. The TES value for a cross-challenge, as compared with the TES values for the compounds in their own tests, gives an indication of the degree to which the two compounds are cross-reactive. The biochemical basis of cross-reactivity has been investigated in some depth by Wulferink et al. (15). Skin sensitization involves the proliferation and release into the circulatory system of clones of T-cells with receptors able to recognize antigenic determinants resulting from covalent binding of reactive compounds to native proteins. When the same sensitizer is encountered again, the protein binding process leads to the same antigenic determinants, and when these are recognized by the circulating T-cells, a sensitization response results. When the sensitizing compound and the challenge compound are different,
a cross-reaction can result in one of three ways. First, when the two compounds modify protein via the same reactive derivative (e.g., BQ could be the reactive species both for itself and for HQ) or by attaching the same group to protein (e.g., 2,4dinitrochlorobenzene and 2,4-dinitrobromobenzene); second, when the two compounds, although attaching different groups to the protein, cause presentation of identical “cryptic peptides” (peptide sequences that in the native protein are concealed within the tertiary structure but become “visible” to T-cell receptors and able to act as antigenic determinants, when the protein is modified); and third, when, although the two compounds produce different antigenic determinants, receptors on some of the T-cell clones are unable to distinguish between them. Consequently, the magnitude of a cross-challenge response depends not only on how much protein binding results from the induction stage and at the challenge stage [probably not to the same extent: the evidence suggests a greater dependence on protein binding at induction than at challenge, for example, ref (16) but on a combination of factors related to the degree of similarity between the various antigenic determinants produced by the two compounds. Cross-challenge studies are nowadays less common than in the past. To illustrate how they can be used, we will discuss the PPD data from the MSIAT studies (Table 2). First, we note that the diagonal entries are the TES values for the compounds in their own tests, that is, when the same compound is used for induction and challenge. These values indicate how strongly the animals were sensitized to the inducing compound. They do not, however, necessarily indicate the relative potencies of the different compounds. This is because different concentrations (based on maximum nonirritant concentration) were used for each compound, at both induction and challenge. For PPD, HQ, BQ, and PAP, the induction concentrations were 0.25, 2, 0.005, and 1%, respectively, and the challenge/cross-challenge concentrations were 2.5, 10, 5, and 5%, respectively. The original reference (12) does not give the concentrations used for OAP and MAP. Considering row 1, PPD induction, we see from the entry in A1 that the animals induced with PPD and challenged with PPD gave a TES of 57%. Thus, the animals were wellsensitized to PPD. When HQ was challenged onto these PPDsensitized animals, the resulting TES (B1) was only 3%, although in its own test (B2) HQ gave a TES of 60%, indicating substantial ability to sensitize and elicit. We can therefore conclude that having been sensitized to PPD made the animals only slightly sensitive to HQ. From the TES value of 33 in C1, we see that having been sensitized to PPD made the animals substantially sensitive to BQ. Similarly, we see from the TES values in D1, E1, and F1 that having been sensitized to PPD made the animals substantially sensitive to OAP, MAP, and PAP. Considering column A, challenges with PPD, we see that animals sensitized to HQ did not respond to PPD challenge (0% in A2), animals sensitized to BQ (57% in its own test, C3) responded only weakly (8% in A3) to PPD, and animals sensitized to OAP, MAP, and PAP responded substantially (11, 27, and 40% in A4, A5, and A6, respectively) as compared to the responses in their own tests (23, 39, and 26% in D4, E5, and F6, respectively). Treating the other entries in Tables 1-3 similarly, it can be seen from the three published sets of test data that: 1. In general, BQ and HQ are strongly cross-reactive with each other. 2. There is some cross-reactivity between BQ/HQ and PPD, but this is much weaker than between BQ and HQ. 3. MAP shows strong sensitization potency, not significantly
Skin Sensitization Chemistry of Pro-Electrophilic Aromatic Compounds Table 3. TES Values for Challenges and Cross-Challenges in the CCET
Chem. Res. Toxicol., Vol. 22, No. 9, 2009 1543 Scheme 1. Radicals and Anions from SH Groups
challenge induction
PPD
HQ
BQ
PPD HQ BQ OAP MAP PAP
19 10 0
5 24 29
0 11 63
3 9
2
OAP
MAP
PAP
8 0 8
3 19 0
0 0 12
Table 4. LLNA Test Data (22) and Calculated Radical Reactivity Parametersa compound
EC3 (%)
PPD HQ BQ OAP MPD MAP MHQ m-dinitrobenzene
0.16 0.11 0.01 0.40 0.49 3.2b 6.0
∆E2 (kcal/mol)
Σσ•
-34.7 -30.1 -30.5 -36.3
1.10 1.03 0.96 1.52
a See Scheme 5 for a basis of ∆E2. Σσ• is the sum of the σ• values for the two substituents in the benzene ring. b The EC3 values in the table were all obtained with acetone/olive oil as the vehicle. A different EC3 value, 0.24%, obtained with DMF as the vehicle, has also been reported for MAP (document SCCP/0978/06; December 19, 2006, link: http://ec.europa. eu/health/ph_risk/committees/04_sccp/sccp_opinions_en.htm#2). A vehicle effect of this magnitude is consistent with the mechanism of abiotic activation by molecular oxygen suggested here, since oxygen solubility and residence times in the oxygen-exposed layers of the stratum corneum are likely to differ between vehicles. For the present analysis, we therefore used the tabulated EC3 values, since they were all obtained with the same vehicle.
weaker than for OAP and para-aminophenol (PAP), and shows significant cross-reactivity with OAP, PAP, PPD, and BQ. In recent decades, information on skin sensitization potential has been acquired with the murine local lymph node assay (LLNA) (17-19), which provides a substantial refinement over the guinea pig assays. The LLNA is a well-established and extensively evaluated approach that underwent significant validation before final acceptance as a method approved by the OECD (Organisation for Economic Cooperation and Development) under test guideline 429 (20). This assay allows for a quantitative assessment of skin sensitization potency through the use of the EC3% value (the EC3 is the percentage concentration of test compound that, when applied to the mouse ear according to the test protocol, produces a 3-fold increase in cell proliferation in the local lymph node) and broader grouping of chemicals as extreme, strong, moderate, weak, and nonsensitizers (21). Recently, LLNA results have been reported for several of these compounds (22). These are summarized in Table 4. It can be seen that in broad agreement with the guinea pig results, BQ is classified as an extreme sensitizer, and HQ and PPD are strong sensitizers in the LLNA (20). The three meta compounds are also skin sensitizers: meta-phenylenediamine (MPD) is classified as strong, whereas MAP and resorcinol (metahydroquinone, MHQ) are both moderate in this classification scheme (21). The implications of all of these animal data are that, first, more than one chemical mechanism is involved [this is implied by the rather weak cross-reactivity between PPD and BQ/HQ and was originally concluded (12) by Basketter and Goodwin]; second, the meta compounds are able to sensitize by some mechanism, which is probably related to those for PPD, HQ, and BQ (this is implied by the observed cross-reactions between
the meta compounds and ortho/para compounds summarized in Tables 2 and 3).
Protein Binding Chemistry The skin sensitization properties of contact allergens can usually be rationalized in terms of their electrophilicity or proelectrophilicity, enabling them to bind to cutaneous proteins. However, it has been suggested that free radical binding mechanisms may also be important (7). For a free radical binding reaction in general, a free radical (Ra•) attacks a substrate (Su) to give another radical (RaSu•) that gives rise, for example by loss or gain of a hydrogen atom, to a stable product (RaSu or RaSuH). Such reaction is more likely to occur if the radical RaSu• is more stable than the radical Ra•, that is, Su acts as a radical trap for Ra•. For a protein binding radical reaction in allergic contact dermatitis (ACD), we can in principle envisage either the protein playing the role of Su, being attacked by a radical derived from the allergen, or the allergen playing the role of Su and being attacked by a radical derived from the protein. There is a third possibility that the binding occurs by reaction between two radicals, one derived from the allergen and one from the protein. This, however, is only really plausible if one of the radicals is stable enough to reach a high concentration in the reaction medium and is unlikely in the situations where the radicals exist as transient intermediates. Sulfhydryl groups of cysteine units in protein are the most likely source of protein-derived radicals, as shown in Scheme 1. Free radical reactions of SH groups in vivo are wellrecognized (23-25). There are several ways in which thiyl radicals can be generated in vivo from cysteine units in proteins and peptidessin general by hydrogen abstraction from S-H by endogenous free radicals (one of the protective roles played by GSH) or by one-electron oxidation of ionised thiyl groups. The formation of thiyl radicals is in particular stimulated by oxidative stress. Cells and tissues can be challenged with oxidative stress from endogenous sources such as normal aerobic respiration and mitochondrial oxygen consumption, oxidative bursts from inflammatory cell phagocytic metabolism, and hydrogen peroxide generated from peroxisome action on fatty acids and other substances (25). Furthermore, normal cellular metabolism and even exercise are all capable of causing oxidative stress-induced injury (25).
PPD and Related Compounds PPD has a variety of reaction pathways potentially available for protein binding leading to skin sensitization (Scheme 2) (26). First, we note that nitrogen atoms and oxygen atoms are very effective at stabilizing free radicals centered on the carbon atoms to which they are bound. Nitrogen is more effective than oxygen, its 2p orbitals being closer in energy to the 2p orbitals of carbon (27). Thus, PPD is readily oxidized, by loss of an electron, to give a radical cation similar to the “Wu¨rster salt” (structure as shown in Scheme 2 but with the hydrogens on the nitrogen atoms replaced with methyl groups) (28-32). The Wu¨rster type salt can bind to protein (i) by attack of a protein-associated
1544
Chem. Res. Toxicol., Vol. 22, No. 9, 2009 Scheme 2. PPD Chemistry
sulfhydryl radical at nitrogen (a reaction product analogous to that shown in Scheme 2 has been observed by Eilstein et al. (26), in this case formed from 2,5-dimethyl-p-benzoquinonediimine), or it can be oxidized further by loss of a hydrogen atom and a proton to give a DI analogous to a quinone (Scheme 2). The DI can then react in a variety of ways, as shown in Scheme 2: (ii) attack of a proteinassociated sulfhydryl radical at nitrogen (giving the same reaction product as from reaction of the Wu¨rster type salt with the sulfhydryl radical); (iii) attack of the sulfhydryl radical at a ring carbon atom in a radical analogue of the Michael addition; (iv) attack of protein-associated thiolate anion at a ring carbon (Michael addition); and (v) hydrolysis to BQ followed by attack of thiolate anion (or other protein associated nucleophile) at carbon in a Michael addition reaction. Here, and elsewhere in this paper, we assign a major role to SH groups in the formation of covalent bonds between the allergen and the cutaneous carrier protein whose modification leads to a cascade of biochemical and cellular processes culminating in the induction or elicitation of skin sensitization. The nature of the carrier protein and its location in the skin are not established, and it is not definitely known that SH groups are the major reaction centers for protein binding of contact allergens. However, SH groups are present in many cytosolic and nuclear proteins, and we see no reason to suppose that they are not present in membrane-bound proteins in skin. Covalent binding of electrophiles to various cytosolic and nuclear proteins via SH groups has been demonstrated (33, 34). We note that since nitrogen is better than oxygen both at stabilizing a radical center and at bearing a positive charge, HQ does not give a radical cation analogous to a Wu¨rster salt. However, semi-
Aptula et al. Scheme 3. MPD Oxidation
quinone radical anions are intermediates in the oxidation of HQ to BQ (28). To the extent that hydrolysis of the DI to BQ occurs, PPD can give the same protein-bound group as HQ or BQ. Thus, the limited cross-reactivity observed (12) can be rationalized. However, it is appropriate to compare BQ and the PPD-derived DI in more detail. Because nitrogen is better than oxygen as a stabilizer of a radical center, the reactions analogous to (ii) and (iii) above are disfavored with BQ relative to DI. In contrast, because oxygen is more electronegative than nitrogen, Michael addition to BQ, reaction v, is favored relative to reaction iv with DI (Scheme 2). Bearing in mind that for PPD but not for HQ there is the possibility of protein binding via the Wu¨rster salt, we consider that overall, PPD is more likely than HQ or BQ to sensitize via free radical pathways and HQ and BQ are more likely than PPD to sensitize via an electrophilic mechanism. These differences can rationalize why PPD is only weakly cross-reactive with HQ and BQ. The above arguments can equally be applied to the ortho analogues of the above compounds. However, we were only able to find data on OAP.
MPD and Related Compounds Although these compounds, unlike their ortho and para counterparts, cannot be oxidized directly to quinones or their nitrogen analogues, they are clearly able to sensitize (12, 22, 35). We can interpret the pattern of sensitization potential in terms of their oxidative hydroxylation or perhydroxylation, leading to quinones or their nitrogen analogues (illustrated in Scheme 3 for oxidation in the 2-position of MPD, 1,3-diaminobenzene; analogous chemistry can apply for MAP and MHQ and for oxidation in the 4-position). In Scheme 3, the process is shown as an abiotic perhydroxylation reaction sequence resulting from attack of molecular oxygen. This corresponds to what Lepoittevin defines as a prehapten mechanism (36). The skin surface and stratum corneum provide a suitably oxygen-rich environ-
Skin Sensitization Chemistry of Pro-Electrophilic Aromatic Compounds Scheme 4. MPDsHypothetical Direct Reaction with Protein-Associated Sulfur Radical
ment, where the test compound is present with a high surface to volume ratio, for nonmetabolic oxidation to occur if the compounds are reactive enough. The initial step in Scheme 3 is the attack of a triplet oxygen molecule on a singlet aromatic molecule. Uncatalyzed reactions between triplet diradicals and singlet unsaturated systems are well-precedentedscycloaddition of triplet carbenes to olefins, for example (37, 38). When triplet oxygen attacks the aromatic ring to produce the diradical intermediates shown, these diradicals are also triplets, that is, both unpaired electrons have the same spin. This means that an intramolecular hydrogen transfer from the ipso carbon to the oxygen cannot occur without prior spin inversion. This spin inversion is likely to be fast [in carbene cycloaddition reactions to olefins the rate of inversion is similar to the rate of rotation about a single bond (39)], making the intramolecular hydrogen transfer electronically possible. However, this H-transfer is likely to be disfavored by the adverse bond angles, and either a sequence of H-abstractions or a pathway involving attack of a second oxygen molecule [analogous to the transient involvement of short-lived pyrosulfonic acids in sulfonation of alkylbenzenes by sulfur trioxide (40)] may be the preferred pathway to an aryl hydroperoxide. Another possibility is that the oxidation proceeds by what Lepoittevin defines (36) as a pro-hapten route, the reaction being accelerated by biological catalytic systems with broad specificity, present in the epidermis (41). The exact nature of the intermediates would then be different from those shown in Scheme 3, but the principle is the same, that the reactivity depends on the ability of the meta substituents to stabilize a free radical intermediate leading to a hydroxylated or perhydroxylated derivative. A different possibility, which we cannot at this point completely exclude, corresponds to what Lepoittevin defines (36) as a direct haptenation route, whereby attack of a proteinassociated sulfhydryl radical on the ring gives an intermediate radical stabilized by two nitrogen or oxygen atoms (Scheme 4). Note that if Scheme 4 applies, MPD and related compounds act as direct haptens, rather than as prohaptens.
Computational Chemistry for the meta-Disubstituted Benzenes To further investigate the above-proposed mechanisms, we carried out molecular orbital computations to quantify the
Chem. Res. Toxicol., Vol. 22, No. 9, 2009 1545 Scheme 5. Basis for Molecular Orbital Calculations of Radical Stability
relative abilities of the substituents to stabilize the free radicals derived from MPD, MAP, and MHQ (resorcinol). The stabilizing effect of the substituents (X and Y in Scheme 5) is independent of the nature of the attacking radical Z, so to simplify the calculations, we based them on attack by the hydrogen radical. We carried out computations for attack of the hydrogen radical on the carbon atom between the two substituents and for attack on the carbon atoms ortho to one substituent and para to the other. All calculations on chemical structure were performed using the Gaussian03 package of programs utilizing the B3LYP/631G(d) level of theory (42). All structures were drawn using the GausView application within Gaussian03 (42), and chemical structures were then optimized using the following criteria: maximum force < 0.000450, rms (root-mean-square) force < 0.000300, maximum displacement < 0.001800, and rms displacement < 0.001200. The formation energy for each of the potential radical species upon addition of a hydrogen radical (Scheme 5) was then calculated (eq 1).
∆Eformation ) Eradical - (Eparent + Ehydrogen radical)
(1)
where ∆Eformation is the formation energy, Eradical is the calculated energy of the parent chemical upon addition of the hydrogen radical, Eparent is the calculated energy of the parent chemical, and Ehydrogen radical is the calculated energy of the hydrogen radical. ∆Eformation is a measure of the ability of the substituents X and Y to stabilize the free radical. The calculations (see third column in Table 4) indicate that the radical from MPD is substantially more stable than the radicals derived from MAP and MHQ (these two being very close in energy). This is in agreement with the σ• values (these are substituent constants, analogous to Hammett constants, for modeling the effects of aromatic substituents on free radical reaction rates) for the OH and NH2 substituents (43). These findings are consistent with MPD being a significantly stronger sensitizer than the other two in the LLNA. The radicals of structure 2 are in all cases slightly more stable than their isomers with structure 1 (Scheme 5). The ∆Eformation values for attack between the two substituents (∆E2) and the Σσ• values are shown in Table 4. We also carried out computations for attack of the hydrogen atom on meta-dinitrobenzene. The results show clearly that the pair of nitro groups stabilize the resulting radical even more strongly than the two NH2 groups of MPD, in agreement with the σ• value of NO2 being greater than that for NH2 (43). This has implications for the protein binding mechanism. If skin sensitization to the meta compounds involves direct reaction with a protein-derived free radical as shown in Scheme 4, then
1546
Chem. Res. Toxicol., Vol. 22, No. 9, 2009
meta-dinitrobenzene would be expected to be a stronger sensitizer than MPD, MAP, or MHQ. However, in guinea pig tests, it is a nonsensitizer (7). We therefore conclude that the direct reaction pathway of Scheme 4 does not apply and can be rejected in favor of the mechanism of Scheme 3, in which the meta-compounds are activated by oxidative conversion to derivatives with three OH and/or amino substituents.
Conclusions The skin sensitization test data on compounds of the general structure C6H4XY where X and Y are hydroxy or amino groups para or ortho to each other, are consistent with two parallel chemical reaction mechanisms: oxidation to electrophilic (protein reactive) quinones, quinone imines, or quinone di-imines or formation of protein reactive free radicals such as the Wu¨rster type radical cation, which can be formed by PPD. The crossreactivity patterns are consistent with HQ reacting mainly by the former mechanism and PPD by the latter mechanism but with some overlap. The ability of the quinone DI derived from PPD to react at a nitrogen atom (26) is another potential source of difference between PPD and BQ/HQ. The sensitization potency of the meta compounds C6H4XY (X and Y hydroxy or amino) can be rationalized in terms of the ability of these groups to stabilize the radical intermediate resulting from attack of a radical species on the carbon atom between the X and the Y positions. Results of quantum mechanics calculations suggest that this radical species is an intermediate in the oxidative hydroxylation of meta-C6H4XY to C6H3XYOH (X, OH, and Y in 1,2,3-orientation), which can then be oxidized to an electrophilic quinone or quinone-imine. Multisubstituted aromatic compounds are among the most difficult to predict, exhibiting a range of possible chemical mechanisms via several potential activation pathways. The research presented here goes some way toward bringing more clarity to this area, but there remains more to be done. For example, the presence of other substituents in addition to the hydroxyl and/or amino substituents can add a further dimension of complexity. Depending on its nature and position in the aromatic ring, a new group may introduce new reaction chemistry possibilities, or it may do no more than modify the rate constants for the reactions that can already take place with the parent diamino-, dihydroxy-, and amino-hydroxy benzenes. Among the new reaction chemistry possibilities that have been proposed are oxidation, when an alkyl group is para to a hydroxyl group, to a Michael electrophilic quinone methide (44), and reaction via an aci/keto tautomer when a nitro group is para to a hydroxyl group (45). It is almost always relatively easy to postrationalize sensitization test results for multisubstituted aromatic compounds, but if predictive capability is to be improved without the use of animal testing, experimental investigation of the reaction chemistry to confirm or modify such rationalizations should be a major priority. Acknowledgment. We acknowledge that the work of Gimenez-Arnau et al., presented at the ERGECD 2005 meeting but, as far as we know, not yet published, whose discussion of Wu¨rster type radical cation chemistry and suggestion that free radical binding mechanisms may also be important and indeed that their importance may have been underestimated, has influenced our thinking. This project was sponsored by Defra through the Sustainable Arable Link Programme, Grant LK0984. The funding of the European Union sixth Framework programme CAESAR Specific Targeted Project (SSPI-022674-
Aptula et al.
CAESAR) and the European Chemicals Agency (EChA) Service Contract No. ECHA/2008/20/ECA.203 is gratefully acknowledged.
References (1) EU (2003) Directive 2003/15/EC of the European Parliament and of the Council of 27 February 2003 amending Council Directive 76/ 768/EEC on the approximation of the laws of the Member States relating to cosmetic products (“7th Amendment to the European Cosmetics Directive”). Off. J. Eur. Union L66, 26-35. (2) Aptula, A. O., and Roberts, D. W. (2006) Mechanistic applicability domains for non-animal based toxicological endpoints. General principles and application to reactive toxicity. Chem. Res. Toxicol. 19, 1097–1105. (3) Roberts, D. W., Aptula, A. O., and Patlewicz, G. (2007) Electrophilic chemistry related to skin sensitization. Reaction mechanistic applicability domain classification for a published data set of 106 chemicals tested in the mouse Local Lymph Node Assay. Chem. Res. Toxicol. 20, 44–60. (4) Roberts, D. W., and Aptula, A. O. (2008) Determinants of skin sensitisation potential. J. Appl. Toxicol. 28, 377–387. (5) Lepoittevin, J.-P., Basketter, D. A., Goosssens, A., and Karlberg, A.T., Eds. (1997) Allergic contact dermatitis. The Molecular Basis, Springer, Heidelberg. (6) Gerberick, G. F., Vassallo, J. D., Foertsch, L. M., Price, B. B., Chaney, J. G., and Lepoittevin, J.-P. (2007) Quantification of chemical peptide reactivity for screening contact allergens: A classification tree model approach. Toxicol. Sci. 97, 417–427. (7) Karlberg, A.-T., Bergstro¨m, M., Bo¨rje, A., Luthman, K., and Nilsson, J. (2008) Allergic contact dermatitissFormation, sructural requirements, and reactivity of skin sensitizers. Chem. Res. Toxicol. 21, 53– 69. (8) Natsch, A., and Gfeller, H. (2008) LC-MS based characterization of the peptide reactivity of chemicals to improve the in Vitro prediction of the skin sensitization potential. Toxicol. Sci. 106, 464–478. (9) Aleksic, M., Thain, E., Roger, D., Saib, O., Davies, M., Li, J., Aptula, A., and Zazzeroni, R. (2009) Reactivity profiling: Covalent modification of single nucleophile peptides for skin sensitisation risk assessment. Toxicol. Sci. 108, 401–411. (10) Enoch, S. J., Madden, J. C., and Cronin, M. T. D. (2008) Identification of mechanisms of toxic action for skin sensitisation, using a SMARTS pattern based approach. SAR QSAR EnViron. Res. 19, 555–578. (11) Landsteiner, K., and Jacobs, J. (1935) Studies on the sensitisation of animals with simple chemical compounds. J. Exp. Med. 61, 643–656. (12) Basketter, D. A., and Goodwin, B. (1988) Investigation of the prohapten concept. Cross reactions between 1,4-substituted benzene derivatives in the guinea pig. Contact Dermatitis 19, 248–253. (13) Aeby, P., Sieber, T., Beck, H., Gerberick, F., and Goebel, C. (2009) Skin sensitization to p-phenylenediamine: The diverging roles of oxidation and N-acetylation for dendritic cell activation and the immune response. J. InVest. Dermatol. 129, 99–109. (14) Wakelin, S. H., Reamer, D., Rycroft, R. J., White, I., and McFadden, J. (1998) Contact dermatitis from paraphenylenediamine used as a skin paint. Contact Dermatitis 39, 92–93. (15) Wulferink, M., Dierkes, S., and Gleichmann, E. (2002) Crosssensitization to haptens: formation of common haptenic metabolites, T cell recognition of cryptic peptides, and true T cell cross-reactivity. Eur. J. Immunol. 32, 1338–1348. (16) Griem, P., Goebel, C., and Scheffler, H. (2003) Proposal for a risk assessment methodology for skin sensitization based on sensitization potency data. Regul. Toxicol. Pharmacol. 38, 269–90. (17) Basketter, D. A., Gerberick, G. F., Kimber, I., and Loveless, S. E. (1996) The local lymph node assay: A viable alternative to currently accepted skin sensitization tests. Food. Chem. Toxicol. 34, 985–997. (18) Basketter, D. A., Evans, P., Fielder, R. J., Gerberick, G. F., Dearman, R. J., and Kimber, I. (2002) Local lymph node assaysValidation, conduct and use in practice. Food Chem. Toxicol. 40, 593–598. (19) Kimber, I., Dearman, R. J., Basketter, D. A., Ryan, C. A., and Gerberick, G, F. (2002) The local lymph node assay: Past, present and future. Contact Dermatitis 47, 315–328. (20) OECD (2002) Organisation for Economic Cooperation and Development. Guidelines for Testing of Chemicals No. 429. Skin Sensitisation: The Local Lymph Node Assay; http://www.oecd.org. (21) Kimber, I., Basketter, D. A., Butler, M., Gamer, A., Garrigue, J. L., Gerberick, G. F., Newsome, C., Steiling, W., and Vohr, H W. (2003) Classification of contact allergens according to potency: Proposals. Food Chem. Toxicol. 41, 1799–1809. (22) Gerberick, G. F., Ryan, C. A., Kern, P. S., Schlatter, H., Dearman, R. J., Kimber, I., Patlewicz, G. Y., and Basketter, D. A. (2005) Compilation of historical local lymph node data for evaluation of skin sensitization alternative methods. Dermatitis 16, 157–202.
Skin Sensitization Chemistry of Pro-Electrophilic Aromatic Compounds (23) Kundu, S. C., and Willson, R. L. (1995) Thiyl (sulfhydryl/thiol) free radical reactions, vitamins, β-carotene, and superoxide dismutase in oxidative stress: design and interpretation of enzymatic studies. Methods Enzymol. 251, 69–86. (24) www.sigmaaldrich.com/life-science/cell-culture/learning-center/mediaexpert/cysteine.html. (25) Baskin. S. I. , and Salem, H. (1997) Oxidants, Antioxidants, and Free Radicals, CRC Press, Boca Raton, FL. (26) Eilstein, J., Gimenez-Arnau, E., Duche, D., Rousset, F., and Lepoittevin, J.-P. (2006) Synthesis and reactivity toward nucleophilic amino acids of 2,5-[13C]-dimethyl-p-benzoquinonediimine. Chem. Res. Toxicol. 19, 1248–1256. (27) Streitweiser, A. (1961) Molecular Orbital Theory for Organic Chemists, Wiley, New York. (28) Gould, E. S. (1959) Mechanism and Structure in Organic Chemistry, p 679, Holt, Rinehart and Winston, New York. (29) Rumpf, P., and Trombe, F. (1938) J. Chim. Phys. 35, 110–113. (30) Rumpf, P., and Trombe, F. (1938) Compt. Rend. 206, 671. (31) Friend, J., and Roberts, N. (1958) A polarographic study of some Wu¨rster salts. Aust. J. Chem. 11, 104–119. (32) Storle, C., and Eyer, P. (1991) Reactions of the Wu¨rster’s blue radical cation with thiols, and some properties of the reaction products. Chem.Biol. Interact. 8, 333–346. (33) Dennehy, M. K., Richards, K. A. M., Wernke, G. R., Shyr, Y., and Liebler, D. C. (2006) Cytosolic and nuclear protein targets of thiolreactive electrophiles. Chem. Res. Toxicol. 19, 20–29. (34) Wong, H. L., and Liebler, D. C. (2008) Mitochondrial protein targets of thiol-reactive electrophiles. Chem. Res. Toxicol. 21, 796–804. (35) Basketter, D., Sanders, D., and Jowsey, I. (2007) The skin sensitization potential of resorcinol: Experience with the local lymph node assay. Contact Dermatitis 56, 196–200. (36) Lepoittevin, J.-P. (2006) Metabolism versus chemical transformation or pro- versus prehaptens? Contact Dermatitis 54, 73–74. (37) Anet, F. A. L., Bader, R. F. W., and Van der Auwera, A. M. (1960) Chemical evidence for a triplet ground state for methylene. J. Am. Chem. Soc. 82, 3217–3218. (38) Frey, H. M. (1960) Chemical evidence for the ground state of methylene. J. Am. Chem. Soc. 82, 5947–5958. (39) Kirmse, W. (1964) Carbene Chemistry, Academic Press, London.
Chem. Res. Toxicol., Vol. 22, No. 9, 2009 1547 (40) Roberts, D. W. (2003) Optimisation of the linear alkyl benzene sulfonation process for surfactant manufacture. Org. Proc. Res. DeV. 7, 172–194. (41) Smith, C. K., and Hotchkiss, S. A. M. (2001) Allergic Contact DermatitissChemical and Metabolic Mechanisms, Taylor & Francis, United Kingdom. (42) Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb, M. A., Cheeseman, J. R., Montgomery, J. A., Jr., Vreven, T., Kudin, K. N. , Burant, J. C., Millam, J. M., Iyengar, S. S., Tomasi, J., Barone, V., Mennucci, B., Cossi, M., Scalmani, G., Rega, N., Petersson, G. A., Nakatsuji, H., Hada, M., Ehara, M., Toyota, K., Fukuda, R., Hasegawa, J., Ishida, M., Nakajima, T., Honda, Y., Kitao, O., Nakai, H., Klene, M., Li, X., Knox, J. E., Hratchian, H. P., Cross, J. B., Bakken, V., Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R. E., Yazyev, O., Austin, A. J., Cammi, R., Pomelli, C., Ochterski, J. W., Ayala, P. Y., Morokuma, K., Voth, G. A., Salvador, P., Dannenberg, J. J., Zakrzewski, V. G., Dapprich, S., Daniels, A. D., Strain, M. C., Farkas, O., Malick, D. K., Rabuck, A. D., Raghavachari, K., Foresman, J. B., Ortiz, J. V., Cui, Q., Baboul, A. G., Clifford, S., Cioslowski, J., Stefanov, B. B., Liu, G., Liashenko, A., Piskorz, P., Komaromi, I., Martin, R. L., Fox, D. J., Keith, T., Al-Laham, M. A., Peng, C. Y., Nanayakkara, A., Challacombe, M., Gill, P. M. W., Johnson, B., Chen, W., Wong, M. W., Gonzalez, C., Pople, J. A. (2004) Gaussian 03, Revision C.02, Gaussian, Inc., Wallingford, CT. (43) Agirbas, H., and Jackson, R. A. (1983) Free radicals in solution. Part 9. Further σ· values from decomposition of substituted dibenzylmercurials. J. Chem. Soc. Perkin 2, 739–742. (44) Bertrand, F., Basketter, D. A., Roberts, D. W., and Lepoittevin, J.-P. (1997) Skin sensitization to eugenol and isoeugenol in mice: Possible metabolic pathways involving ortho-quinone and quinone methide intermediates. Chem. Res. Toxicol. 10, 335–343. (45) Roberts, D. W., Patlewicz, G., Dimitrov, S. D., Low, L. K., Aptula, A. O., Kern, P. S., Dimitrova, G. D., Comber, M. I., Phillips, R. D., Niemela¨, J., Madsen, C., Wedebye, E. B., Bailey, P. T., and Mekenyan, O. G. (2007) TIMES-SSsA mechanistic evaluation of an external validation study using reaction chemistry principles. Chem. Res. Toxicol. 20, 1321–1330.
TX9000336
