Abstract
The vehicle in which an allergen is presented to the skin has been recognized to have an effect on the skin-sensitizing potency of the allergen. Typical vehicles used to evaluate the skin sensitization potential of fragrance materials include ethanol, diethyl phthalate, or a combination of the two. The authors conducted a series of studies to evaluate each of these vehicles for their utility in the murine local lymph node assay and to investigate the potential differences in skin sensitization resulting from their use. Four fragrance materials were tested in four different vehicles. The test materials were p-t-butyl-α-methylhydrocinnamic aldehyde, geraniol, eugenol, and hydroxycitronellal. The vehicles were diethyl phthalate, 1:3 ethanol:diethyl phthalate, 3:1 ethanol:diethyl phthalate, and ethanol. Each of the fragrance materials was tested at five dose levels ranging from 0.3% to 50% w/v. In all four vehicles, each material tested elicited positive responses, exhibiting weak to moderate skin sensitization potential. Overall, p-t-butyl-α-methylhydrocinnamic aldehyde exhibited the most potency, followed by eugenol, geraniol, and hydroxycitronellal. The sensitization potential of both p-t-butyl-α-methylhydrocinnamic aldehyde and geraniol was greatest when the vehicle was ethanol. The sensitization potential of eugenol was greatest in 3:1 ethanol:diethyl phthalate, but the sensitization potential for hydroxycitronellal was greatest in 1:3 ethanol:diethyl phthalate. The strength of the sensitization response was observed to vary with the vehicle; however, the results did not show any clear pattern of one vehicle over another regarding skin sensitization.
The Local Lymph Node Assay (LLNA) is an alternative to traditional animal skin sensitization models for the identification of potential contact allergens and has undergone rigorous development, evaluation, and validation (Kimber and Dearman 1991; Kimber and Basketter 1992; Kimber et al. 1994; Dearman, Basketter, and Kimber 1999; Gerberick et al. 2000). It has been endorsed as a stand-alone method for skin sensitization testing by the US Interagency Coordinating Committee on the Validation of Alternative Methods (ICCVAM) and the European Centre for the Validation of Alternative Methods (ECVAM) (National Institutes of Health [NIH] 1999; ECVAM 2000). More recently, the Organization for Economic Co-Operation and Development (OECD) adopted an updated guideline for the LLNA (OECD 2002). Activity in the LLNA is measured as a function of the level of induced T-lymphocyte proliferation (a necessary component of contact sensitization) in the lymph nodes draining the site of topical chemical application. The strength and magnitude of the lymph cell proliferative response was found to closely correlate with the extent to which sensitization will occur and is measured by the amount of radiolabeled thymidine incorporated into the dividing cells (Kimber and Dearman 1991). In addition to its application as a method to identify potential contact allergens, the LLNA presents the opportunity for the objective and quantitative measure of relative skin sensitizing potency (Basketter et al. 1999a, 2000).
It has been recognized for some time that the vehicle matrix in which a chemical allergen is presented to the skin can have a profound effect on the sensitizing potency of the allergen (Kligman 1966). Formal evaluations were previously undertaken to assess the impact of the recommended vehicles for use in the LLNA on sensitization potential (Lea et al. 1999; Warbrick et al. 1999). In addition to the recommended vehicles, the currently adopted guidelines for the LLNA provide for the use of alternative vehicles for which there is sufficient scientific rationale (OECD 2002). In the study reported here four alternative vehicles, diethyl phthalate (DEP), 1:3 ethanol:diethyl phthalate (1:3 EtOH:DEP), 3:1 ethanol:diethyl phthalate (3:1 EtOH:DEP), and ethanol (EtOH), were evaluated for their utility in the LLNA and influence on the skin sensitization potential of fragrance materials. The vehicles were selected because most fragrance materials are soluble in EtOH and DEP. In addition, the combination of EtOH and DEP is representative of the matrix in which human skin exposure to commercial fragrance materials is likely to occur and is employed at the Research Institute for Fragrance Materials (RIFM) for both hazard and safety assessments (Api 2002). A combination is preferred because under occlusive condition, such as the human repeated-insult patch test, irritation is increased with increasing levels of EtOH and although a rare occurrence, EtOH alone can induce sensitization in humans (Stotts and Ely 1977). On these bases, the vehicles in the present study were chosen to reflect increasing concentrations of EtOH.
The fragrance materials selected were p-t-butyl-α-methylhydrocinnamic aldehyde (BMHCA), geraniol, eugenol, and hydroxycitronellal. Each material represents a different structural class of commonly used fragrance ingredients, which are freely soluble in the selected vehicles (Arctander 1969). Analyses of the chemical structure of BMHCA, eugenol, and hydroxycitronellal reveal structural alerts for potential toxic dermal effects such as sensitization (Ford et al. 2000). Though geraniol does not possess a structural alert for topical effects, it has been postulated that geraniol may be converted to geranial, which possesses such an alert, in situ by epidermal alcohol dehydrogenases (Smith and Hotchkiss 2001). Sensitization reactions to geraniol have been observed in both human and animal studies (Klecak, Geleick, and Frey 1977; Marzulli and Maibach 1980). Overall, experience has shown each material to be a mild to moderate sensitizer in both human and animal models of sensitization. Such mild to moderate sensitizers are suitable choices for assays to investigate the effect of vehicle on sensitization potential.
MATERIALS AND METHODS
Animals
Young adult (8 to 12 weeks old) CBA/Ca strain male mice (Harlan Interfauna UK, Shaw’s Farm, Blackthorne, Bicester, Oxon, UK) were used throughout the study. Mice were housed in groups of four per cage under standard conditions. Fluorescent light with a 12-h light and 12-h dark cycle illuminated the room. Temperature and relative humidity were maintained at 19°C to 25°C and 30% to 70%, respectively. Food (Special Diets Services Ltd [SDS] Porton Combined Diet [PCD] pelleted diet; Special Diets Services Ltd., Witham, United Kingdom) and water were available ad libitum.
The study described herein was performed in accordance with the Animals (Scientific Procedures) Act 1986 as detailed in the guidance provided by the Home Office (2000).
Test Material
The test substance BMHCA (purity 98.6%) was supplied by Givaudan (Switzerland). Geraniol (purity 98.5%) and hydroxycitronellal (purity 98.7%) were obtained from Bush Boake Allen Limited (UK). Eugenol (purity 99.9%) was supplied by Haarmann & Reimer GmbH (Germany). The following vehicles were supplied by Bush Boake Allen Limited (UK): DEP, 1:3 EtOH:DEP, 3:1 EtOH:DEP, and EtOH. The structures for all test materials are given in Figures 1 and 2.
Local Lymph Node Assay
The assays were conducted according to the method of Kimber et al. (1992, 1994). Groups of male mice (n = 4) were dosed topically on the dorsum of both ears with 25 μl of test material, at one of five concentrations, in one of the four vehicles or to the same volume of the vehicle alone, which acted as the control (see Table 1). Dosing occurred daily for 3 consecutive days. The animals “rested” for 2 days and on the sixth day after the first application, all mice were injected intravenously by the tail vein with 250 μl of phosphate-buffered saline (PSB) containing 20 μCi of [3H]methyl thymidine (3HTdR; specific activity 2.0 Ci/mmol; Amersham Pharmacia Biotech UK Limited, UK). Five hours later, the mice were euthanized and the draining auricular lymph nodes were excised and pooled for each experimental group. Suspensions of the lymph node cells were prepared by mechanical disaggregation through 200-mesh stainless steel gauze. The cell suspensions were washed three times with PBS and precipitated overnight at 4°C with 5% w/v trichloroacetic acid (TCA). The samples were then pelleted by centrifugation. The cells were resuspended in 1ml of TCA and transferred to scintillation vials containing 10 ml of scintillation fluid (Optiphase MP, LKB). The incorporation of 3HTdR was measured by β-scintillation counting and expressed as counts per minute (cpm) per lymph node for each experimental group. For each concentration of test material, a stimulation index (SI) relative to the concurrent vehicle-treated control was calculated. The SI value for each test material was calculated by dividing the mean cpm at a given dose level by the mean cpm of the vehicle control group. A material was considered a sensitizer if at least one concentration of the test material was observed to have an SI value of 3 or more.
Mathematical Analysis
The EC3 value, or estimated concentration of test material required to elicit an
SI of 3 or more, was derived from the dose-response data by linear interpolation.
This value was taken as a measure of relative sensitization potential for each
material. Using two data points on the dose response curve, one immediately above
and one below the SI value of three, the EC3 value was calculated utilizing the
following equation presented by Basketter
et al. (1999b):
Additionally, SI values were compared, separately for each material, using an analysis of variance fitting: (1) a factor to represent differences in overall SI values between vehicles and (2) linear and quadratic dose covariates representing a quadratic regression between SI values and dose (Snedecor and Cochran 1980). The quadratic regression was allowed to differ in each vehicle. Statistical comparisons of the linear (representing the steepness of the response) and quadratic (representing curvature of the response) components were made between pairs of vehicles.
RESULTS
The influence of each vehicle on the response to the four fragrance materials in each LLNA was evaluated. The LLNA dose-response data obtained with each fragrance material and each vehicle tested are presented along with the calculated SI values in Table 1. The calculated EC3 values are given in Table 2. The assays conducted with BMHCA utilizing EtOH or DEP as the vehicle generated low EC3 values when compared to the maximum dose of 50%. This was not the case with eugenol, geraniol, and hydroxycitronellal. Therefore, a lower dosing range was selected for the two assays on BMHCA in the combined vehicle (EtOH:DEP) in order to more accurately calculate the EC3 values.
The levels of thymidine incorporation measured for the vehicle treated control animals extended over a relatively narrow range, from a low of 133 cpm node−1 for EtOH to a high of 287 cpm node−1 for 1:3 EtOH:DEP. Regardless of vehicle, all test materials resulted in a dose-dependent induction of local lymph node cell (LNC) proliferation and SI values of greater than 3 in one or more of the concentrations tested, therefore each can be regarded as potential skin sensitizers.
Graphs of the dose-response data based on concentration of test material to concentration of EtOH in the vehicle are presented in Figure 3. These contour plots illustrate graphically general trends of LNC proliferation for each material; however, it should be cautioned that such contour curves are based on interpolation and/or extrapolation of the limited individual data sets for each material. The dose responses of eugenol and BMHCA show a trend toward increasing LNC proliferation as both test material and EtOH concentrations increase (Figure 3A and B ). Eugenol achieved the highest maximum SI value of 18.7 in 3:1 EtOH:DEP and the lowest maximum in DEP at 6.4. BMHCA at the highest concentration tested (50%) in DEP and EtOH achieved maximal SI values of 9.2 and 15.67, respectively. The dose response to hydroxycitronellal shows test material concentration dependence only, with a slight boost in proliferation at the highest EtOH concentration (Figure 3C ). The SI values obtained at the highest concentration (50%) of hydroxycitronellal varied over a narrow range of 5.3 to 7.2. The response in geraniol shows no clear trend (Figure 3D ). The two highest SI values of 11.6 and 7.8 were obtained with EtOH and DEP, respectively.
Analysis of variance was used to assess the statistical significance of the steepness and curvature of the dose responses obtained with the use of each vehicle with individual materials. For BMHCA, statistically significant differences (p <.05) were observed in the dose response between EtOH and DEP alone and the combined vehicles. There also tended to be increasing responses at higher dose levels leading to greater curvature when BMHCA was tested in EtOH. The dose response obtained with geraniol in EtOH was statistically significantly steeper than that observed for the other vehicles. At higher doses, significant differences were observed in the curvature of the dose responses to eugenol. The response to hydroxycitronellal in EtOH showed statistically significant differences (steeper and more curved) as compared to the other vehicles.
Comparisons of the derived EC3 values (Table 2) indicate that the vehicle matrix in which skin exposure occurs influences the relative sensitization potential of each material. The sensitization potential of both BMHCA and geraniol was greatest when the vehicle was EtOH resulting in EC3 values of 3.0% and 5.6%, respectively. The sensitization potential of eugenol was greatest in 3:1 EtOH:DEP with an EC3 value of 5.3%. In each of these three materials there was at least a threefold difference between the highest and lowest EC3 values resulting from the assays conducted in the different vehicles. The sensitization potential for hydroxycitronellal was greatest in 1:3 EtOH:DEP. However, hydroxycitronellal exhibited only modest differences in EC3 values obtained from the use of each vehicle.
When EC3 values were used to rank the test materials for their potential to induce skin
sensitization in each vehicle, the results were as follows: EtOH: BMHCA > geraniol > eugenol > hydroxycitronellal DEP: BMHCA > geraniol > eugenol > hydroxycitronellal 1:3 EtOH:DEP: eugenol > BMHCA > hydroxycitronellal > geraniol 3:1 EtOH:DEP: eugenol > BMHCA > hydroxycitronellal > geraniol BMHCA: EtOH > DEP > 3:1 EtOH:DEP > 1:3 EtOH: DEP Eugenol: 3:1 EtOH:DEP > 1:3 EtOH:DEP > EtOH > DEP Geraniol: EtOH > DEP > 1:3 EtOH:DEP > 3:1 EtOH:DEP hydroxycitronellal: 1:3 EtOH:DEP > DEP > 3:1 EtOH:DEP > EtOH
When the EC3 values were used to rank the vehicles according to their potential
to increase the skin sensitization potential of each test material, the results were as
follows:
DISCUSSION
The selection of an appropriate vehicle for use in hazard identification and risk assessment of dermal effects is an important consideration, as the choice of vehicle is known to have an effect on the skin sensitization potential of a material (Lea et al. 1999; Warbrick et al. 1999). The greatest exposure to fragrance materials results from alcohol-based products on the skin (Cadby, Troy, and Vey 2002). Therefore, EtOH and EtOH-containing vehicle systems represent a realistic vehicle for the investigation of dermal effects to fragrance materials. The current investigation was conducted with two aims. First was to use known sensitizers to examine whether the selected EtOH vehicles were appropriate for use in the LLNA—that is, whether the materials tested could be identified as sensitizers in the chosen vehicles. Second was to examine if a consistent vehicle effect on sensitization potential could be established over the range of EtOH concentrations selected.
Based on the outcomes of the LLNAs reported here, regardless of the nature of the vehicle employed, all the test materials were observed to have the potential to cause skin sensitization. However, the magnitude of the sensitization response varied with the vehicle. Utilizing the potency estimations based on EC3 values of Basketter et al. (2000) (which segregates potency into five classes, with Human Class 1 containing the strongest allergens and Human Class 5 containing nonsensitizers), BMHCA, eugenol and geraniol would be classified as Human Class 2 (moderate) to Class 3 (weak) for the currently reported series of assays. Hydroxycitronellal would be classified as Human Class 3 (weak). The observed potencies are similar to those reported for each material in the recommended LLNA vehicles (Basketter et al. 2000). These results show that the EtOH and/or DEP vehicle system is appropriate for use in the LLNA and is a suitable alternative to the recommended vehicles.
By comparing the dose-response (Figure 3) and calculated EC3 values, there is no clear pattern of one vehicle over another regarding the potential to induce skin sensitization. A consistent pattern was not observed based on concentration of EtOH or DEP. In the case of BMHCA, eugenol, and geraniol, EtOH and 3:1 EtOH:DEP appeared to have more of an effect inducing skin sensitization than the other vehicles. However, DEP alone was observed to be the second most potent vehicle for eugenol and geraniol. Furthermore, hydroxycitronellal exhibited the most potency in the low EtOH vehicle and DEP alone. Similar difficulties in making generalizations about the effects of vehicle on potency estimations have been highlighted previously following an investigation with the recommended vehicles utilized in the LLNA (Wright et al. 2001).
There are a number of relevant mechanisms by which the vehicle system may influence skin sensitization potential. The observed differences in EC3 values reported here could be attributed to vehicle-dependent modifications of skin absorption, skin metabolism, bioavailability, irritation, and/or the vehicle may effect the expression of epidermal cytokine and migration of Langerhans cells (Kligman 1966; Dearman et al. 1996; Heylings et al. 1996). Available skin absorption data for hydroxycitronellal and BMHCA in the EtOH vehicle system allowed for comparison of skin absorption to results in both human and murine sensitization tests. Such comparison, given below, further illustrates the complex mechanisms by which a vehicle may alter potency and the difficulty of making generalizations. Additionally, the comparisons show that this becomes most evident when extrapolating between species.
Skin absorption studies in rats (Tonge 1995) showed that after 72 h there was slightly more absorption of 20% hydroxycitronellal in 3:1 EtOH:DEP than in EtOH alone. The results from the LLNA presented here show a slightly greater sensitization potential with the 3:1 EtOH:DEP vehicle than with EtOH alone. This may be due to differences in the amount of hydroxycitronellal absorbed through the skin. Additionally, in human repeated insult patch tests (Letizia and Api 2001), it was shown that hydroxycitronellal in 3:1 EtOH:DEP had greater sensitization potential than hydroxycitronellal in DEP alone, just the opposite of the outcome in the LLNA reported here.
In a human skin absorption study (RIFM1994) using BMHCA in 70% EtOH, very little BMHCA was absorbed through the skin, but in this current study, BMHCA in EtOH was shown to have the greatest potential for skin sensitization. In addition, human repeated-insult patch test data have shown that in a high-EtOH vehicle (3:1 EtOH:DEP), BMHCA produced sensitization reactions; however, in a low-EtOH vehicle (1:3 EtOH:DEP), no reactions were observed (Cocchiara and Api 2003).
In summary, the data presented here show that the EtOH-and-DEP vehicle system is suitable for use in the local lymph node assay when assessing the sensitization potential of fragrance materials. The strength of the sensitization response was observed to vary with the vehicle; however, the results did not show any clear pattern of one vehicle over another regarding the potential to induce skin sensitization. The LLNA has been increasingly used to identify both the hazard and sensitization potential of fragrance materials prior to conducting studies in humans. The results presented here and by others indicate that vehicle-dependent variations in the LNC response result in differing sensitization potentials. Therefore, it is appropriate to design studies utilizing the vehicle that is most relevant to human exposure. The greatest exposure to fragrance materials results from hydroalcoholic type products on the skin, as such, EtOH represents a realistic vehicle for use in their assessment. Due to the rare occurrence of sensitization and the irritation that can result from high concentrations of EtOH, a combination of EtOH and DEP is preferred. This combination is equally realistic, as DEP is present in many fragrance compounds as both a solvent and fluidizer. On these bases, EtOH and DEP are routinely used as vehicles in human studies conducted on fragrance materials. By utilizing the EtOH-and-DEP vehicle system in the LLNA, the observed sensitization potential can be extrapolated to humans with more confidence. This increased confidence can then be applied to aid in the design of human studies that utilize EtOH and DEP as a vehicle. In addition, the results obtained in the LLNA may be more readily compared to the large number of studies already conducted on fragrance materials in humans utilizing this same vehicle.
Footnotes
Figures and Tables
This article is based in part on work presented in poster format at the Annual Meeting of the American College of Toxicology, 2000. The local lymph node assays were conducted at the Central Toxicology Laboratory (CTL), Macclesfiled, United Kingdom.