Alternative Methods for Skin Sensitisation Testing

The Report and Recommendations of ECVAM Workshop 191,2

Reprinted with minor amendments from ATLA 24, 683-705

Odile de Silva3, David A. Basketter4, Martin D. Barratt4, Emanuela Corsini5, Mark T.D. Cronin6, Pranab K. Das7, Joachim Degwert8 Alexander Enk9, Jean Luc Garrigue3, Conrad Hauser10, Ian Kimber11, Jean-Pierre Lepoittevin12, Josette Peguet13, and Maria Ponec14
3L'Oréal, 1 Avenue Eugéne Schueller, 93600 Aulnay-sous-Bois, France; 4Unilever Environmental Safety Laboratory, Colworth House, Sharnbrook, Bedford MK44 1LQ, UK; 5Laboratoire de Toxicologie, Istituto di Scienze Farmacologiche, Via Balzaretti 9, 20133 Milan, Italy; 6School of Pharmacy and Chemistry, Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, UK; 7Department of Dermatology and Pathology, Academic Medical Center, University of Amsterdam, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands; 8Beiersdorf Immunology, Cosmed Division, PGU Skin Research Center, Unnastrasse 48, 20245 Hamburg, Germany; 9Department of Dermatology, University of Mainz, Langenbeckstrasse 1, 55131 Mainz, Germany; 10Allergy Unit, Division of Immunology and Allergy, Clinique de Dermatologie, Hôpital Cantonal Universitaire, 1211 Geneva 14, Switzerland; 11ZENECA Central Toxicology Laboratory, Alderley Park, Macclesfield, Cheshire SK10 4TJ, UK; 12Laboratoire de Dermatochimie, Clinique Dermatologique, CHU, 67091 Strasbourg, France; 13INSERM UR 346, Clinique Dermatologique, Hôpital Edouard Herriot, 69437 Lyon 03, France; 14Department of Dermatology, University Hospital Leiden, 2300 RC Leiden, The Netherlands

1ECVAM - The European Centre for the Validation of Alternative Methods. 2This document represents the agreed report of the participants as individual scientists.

Address for correspondence: Dr O. de Silva, L'Oréal, 1 Avenue Eugéne Schueller, 93600 Aulnay-sous-Bois, France

Address for reprints: ECVAM, TP 580, JRC Environment Institute, 21020 Ispra (VA), Italy

Preface

This is the report of the nineteenth of a series of workshops organised by the European Centre for the Validation of Alternative Methods (ECVAM). ECVAM's main goal, as defined in 1993 by its Scientific Advisory Committee, is to promote the scientific and regulatory acceptance of alternative methods which are of importance to the biosciences and which reduce, refine or replace the use of laboratory animals. One of the first priorities set by ECVAM was the implementation of procedures which would enable it to become well-informed about the state-of-the-art of non-animal test development and validation, and the potential for the possible incorporation of alternative tests into regulatory procedures. It was decided that this would be best achieved by the organisation of ECVAM workshops on specific topics, at which small groups of invited experts would review the current status of various types of in vitro tests and their potential uses, and make recommendations about the best ways forward (1).

The workshop on Alternative Methods for Skin Sensitisation Testing was held in Angera, Italy, on 24-28 April 1995, under the co-chairmanship of Odile de Silva (L'Oréal, Aulnay-sous-Bois, France) and David Basketter (Unilever Environmental Safety Laboratory, Bedford, UK). The objective of the workshop was to prepare a state-of-the-art review on non-animal methods for skin sensitisation testing, and to formulate recommendations for making progress toward the total replacement of the current animal tests.

Introduction

Novel approaches which permit both the refinement and reduction of animal use for the identification of skin sensitisers are now being employed. Our understanding of the chemistry and biology of skin sensitisation has progressed to the point where consideration of the development of possible replacement alternative methods is appropriate. The workshop participants reviewed state-of-the-art methodologies for skin sensitisation testing, including the use of (quantitative) structure-activity relationships ([Q]SAR) and expert systems, and the immunology and cell biology underlying the phenomenon of skin sensitisation, to determine where the most realistic opportunities for the successful development of non-animal alternative methods lie.

The in vitro methods available, which principally involve the use of cell monolayers, co-cultures or isolated skin explants, were evaluated, with emphasis given to those which appear to be most promising in terms of the development of predictive assays within a reasonable time scale. Other fields of research which could lead to the development of more complex test systems (able to more accurately mimic the events which lead to skin sensitisation in vivo) were also discussed at the workshop. This report concludes with clear recommendations for future initiatives which are required if progress is to be made with respect to the replacement of animal procedures for skin sensitisation testing.

Major Biological Events in Skin Sensitisation

Skin sensitisation is the process following the epicutaneous application of a substance to the skin which results in an immunological response specific for this substance. Skin sensitisation is also called "delayed contact hypersensitivity", "contact hypersensitivity", "contact allergy" or "allergic contact dermatitis".

The induction phase of skin sensitisation occurs when the substance is applied to the skin of a naive individual, resulting in a state of specific hypersensitivity; it does not result in any clinical symptoms (unless the substance is also an irritant). The main cells involved in the induction phase are epidermal Langerhans cells (and other cutaneous dendritic cells), keratinocytes and T-lymphocytes.

The elicitation, or challenge, phase occurs after a subsequent encounter of the sensitised individual with the inducing substance, usually after a certain period of time (days or months). This phase results in the clinical manifestations of erythema and/or oedema at the site of contact. These manifestations occur 24-48 hours after the actual exposure (that is, "delayed" hypersensitivity).

To behave as a contact allergen, a chemical must penetrate into the skin and react with proteins. Whilst our knowledge of this process is incomplete, it is nevertheless necessary to take it into consideration when developing methods to predict skin sensitisation.

Sensitisers

With regard to primary and secondary in vitro T-cell responses, the best studied group of sensitisers are electrophilic haptens (for example, trinitrobenzene sulphonic acid [TNBS]). Studies undertaken in both mice and humans have been published (2, 3). There is little information about other sensitisers. TNBS, which is water-soluble, seems to generate T-cell antigens in antigen-presenting cells (APC) which can then be recognised by T-cells sensitised to the hydrophobic allergen trinitrochlorobenzene (TNCB). In turn, T-cells from naive mice incubated with TNBS-modified APC and then adoptively transferred into other naive mice can confer contact hypersensitivity to TNCB (4), suggesting that at least some of the T-cells which were activated by TNBS-modified APC are able to recognise T-cell antigens generated in vivo by the epicutaneous application of TNCB.

From these observations it can be concluded that at least some lipophilic haptens and their water-soluble analogues are able to generate cross-reacting T-cell antigens. As water-soluble haptens are required for in vitro modification of APC, hydrophilic analogues have been used in some cases. These may generate antigens which cross-react with antigens generated by their lipophilic counterparts, although this may not always be the case, because lipophilic haptens and their water-soluble counterparts can attack different regions in the same protein.

Covalent coupling of hapten to amino acid side chains is one of the critical steps in generating the T-cell antigen. The principles of organic chemistry may help in predicting those analogues which may couple to the same amino acids. The coupling properties of haptens and their analogues to proteins can often be verified experimentally. Alternatively, some lipophilic compounds can first be dissolved in apolar diluents, such as dimethylsulphoxide, before adding them to the cell suspension. Chemicals which are unstable in water may pose a particular problem in this context. Another point to consider with regard to sensitisers is the fact that many sensitising compounds are prohaptens, and their active metabolites (that is, the haptens which couple to proteins) are unknown. For some of these compounds, active metabolites can be predicted (5). In addition, it should be emphasised that these considerations may be valid for electrophilic haptens (2), but it is controversial whether nucleophilic haptens exist, and the way in which ambiphilic haptens are presented by APC is unknown (6, 7).

Carriers

A carrier is usually a protein which permits a hapten to be recognised by T-cells. The carriers of in vitro hapten-modified APC, and those in sensitised mice in vivo, are unknown, but they are likely to be self-proteins. In order to function as a carrier, the hapten-modified self-protein must enter either the exogenous or endogenous pathways of antigen processing. Subsequently, self-protein-derived peptides to which the hapten is covalently linked bind to the peptide-binding groove of the major histocompatibility complex (MHC) class I and/or MHC class II molecules. The factors influencing the coupling reaction which the hapten undergoes during antigen processing are presently unknown. Genetic heterogeneity determining the rules of antigen processing and presentation may not be relevant with a homogeneous genetic background, such as inbred mouse strains, but it may play a major role in an outbred population, such as in humans.

The MHC hapten-peptide complex is recognised by antigen-specific receptors on T-cells. The structural requirements for peptide binding are known for some mouse MHC haplotypes, and for some of the human MHC proteins. Nevertheless, information on the mouse and on humans is far from complete. Knowledge about the structural requirements for T-cell recognition of hapten is limited to that for trinitrophenyl (TNP) group-bearing peptides in limited number of mouse haplotypes (6, 8). At present, no information has been published for humans. However, it has become clear from studies conducted with TNP that many proteins may potentially act as carriers; statistically, the structural requirements (for example, the position and physicochemical properties of the MHC-anchoring amino acid side chains, and the position of the hapten acceptor group) may enable numerous proteins to function as carriers. With regard to hapten-carrier conjugates, the molecular requirements that precede binding to MHC are largely unknown, and these could limit the number of carriers. Alternatively, the hapten could bind to peptides already embedded in the MHC molecule.

At present, it is unknown whether in vivo processing of hapten-carrier conjugates is required for sensitisation and/or challenge reactions. T-cell recognition has a threshold with regard to the number of MHC peptide complexes on the APC, but it is unclear whether this is a critical issue with regard to T-cell recognition of antigens generated from the carrier and hapten. If a large number of proteins can function as carriers, it may not represent a critical issue. As haptens may couple directly to MHC proteins, it is possible that antigens generated in this way could play a role in sensitisation, although experimental evidence for the existence of this mechanism is lacking.

Antigen-Presenting Cells and Other Types of Skin Cells

APC and other cells found in the skin have been shown to play an important role in the induction phase of sensitisation, as well as in the early phases of elicitation. Epidermal Langerhans cells (LC) and other dendritic cells (DC) in the dermis pick up hapten and migrate to the draining lymph nodes, where they interact with T-lymphocytes, thereby resulting in the proliferation and clonal expansion of specific T-cell sub-populations.

Epidermal LC exhibit the earliest measurable responses following the application of a sensitising hapten. Interleukin (IL)-1β mRNA signals increase as early as 15 minutes after hapten application (9). The second cytokine for which there is an increase in mRNA within the epidermis is tumour necrosis factor (TNF)-a, a product of keratinocytes. There is good evidence that IL-1β is one of the signals for the induction of TNF-a following the application of sensitisers such as TNCB and dinitrofluorobenzene (DNFB [10]). The secondary induction of other cytokine mRNAs can be observed subsequently. Activation of the cytokine cascade by LC-derived IL-1β is also the initial signal for morphological changes of the LC within the epidermis, the upregulation of the expression of MHC class II molecules on their surfaces, and the subsequent decrease in their density in the epidermis.

IL-1β and TNF-a seem to play critical roles in the migration of LC and other DC from the skin to the draining lymph nodes (11, 12). The upregulation of MHC class II expression is accompanied by an increase in accessory cell function in a culture model of LC activation (13). This has been shown to happen in vivo after hapten application, thereby enabling LC to activate naive T-cells (14). The individual molecules involved in this upregulation of accessory function include intracellular adhesion molecule ICAM-1, ICAM-3, and B7 molecules (B7.1 and B7.2 [15, 16]).

Induction of the cytokine cascade induced by hapten application may have various other consequences. Pro-inflammatory cytokines may prepare the endothelial cells for the extravasation of circulating leukocytes. In fact, it has been shown that hapten application may induce adhesion molecules such as ICAM-1, E-selectin, and vascular adhesion molecule (VCAM)-1 (15, 17, 18). T-cells arriving in the skin can interact with APC and further amplify hapten-driven skin inflammation.

T-cells

T-cells have been shown to transfer delayed contact hypersensitivity. The specificity of the reaction is mediated by the T-cell receptor for antigen. This is also the case for the recognition of hapten. Although there is good evidence that T-cells with aβ-type T-cell receptors are involved in sensitisation and elicitation reactions, T-cells with ysigma-type T-cell receptors may also be involved (19). T-cells expressing CD4 and CD8 have been cloned and utilised in studies of hapten-peptide recognition.

Hapten-specific CD4+ and CD8+ T-cells can mediate contact hypersensitivity (4, 20). The T-cell receptors of some TNP-specific T-cells have been cloned and sequenced. Although evolution has shaped the T-cell repertoire for the recognition of infectious agents, it seems that, in the case of TNP, a wide variety of receptors are able to recognise the compound when it is present on APC (6). This is not surprising, because the T-cell repertoire, which has an estimated potential number of 1015 diverse receptors, should be able to recognise any haptenic substitutions of proteins. Nevertheless, the existence of a large variety of T-cell receptors which are able to recognise other sensitisers has yet to be demonstrated.

Not all T-cell types mediate contact sensitivity. It has become clear that the type of T-cell effector function is closely associated with the accompanying profile of soluble hormone-like mediators, referred to as cytokines (21). In addition, cytolytic machinery and the presence of specific cell-surface receptors determine the effector functions of T-cells.

The conditions of previous (including primary) activation appear to determine the set of effector functions which occur in response to restimulation. For example, in many experimental systems it has been shown that when T-cells with a type 1 cytokine pattern (IL-2, interferon [IFN]-g) were induced, delayed hypersensitivity developed, but when T-cells with a type 2 cytokine profile (IL-4, IL-5) predominated, no delayed hypersensitivity was elicited, although other parameters of an adaptive immune response (for example, antibody production) were found (21). Thus, it is questionable whether proliferation, which is the most frequently used in vitro measurement for T-cells, is relevant with regard to effector function. Nevertheless, it has been demonstrated that T-cell proliferation in vitro correlates well with the production of contact hypersensitivity in vivo.

In fact, clonal expansion of T-cells is associated with all types of T-cell-dependent immune responses. Clonal expansion of T-cells towards antigen application thus provides information about the immunogenicities of antigens, but not about their capacities to induce delayed hypersensitivity. Consequently, T-cell proliferation in vitro provides information on the T-cell recognition of antigen, but not necessarily on their capacities to induce delayed hypersensitivity.

In conclusion, the induction of contact hypersensitivity involves a complex series of cellular and molecular events. The relative contributions and interdependence of these events are not completely understood. It remains to be demonstrated whether an in vitro test, or several in vitro tests in combination which model the critical steps in sensitisation, can replace animal experiments for predicting contact allergic reactions in humans.

Current Procedures for Skin Sensitisation Testing

The important requirements in the context of this report are those tests which must be conducted in the European Union (EU) for the safety evaluation/risk assessment, in a regulatory context, of chemicals and products with respect to their skin sensitisation potentials. This subject has been reviewed previously by Botham et al. (22), but their review paid only limited attention to the status of alternative methods and their validation.

Regulatory Tests and Their Interpretation

In the EU, two guinea pig test methods are described in Annex V to the Dangerous Substances Directive, Directive 67/548/EEC (23). These are the guinea pig maximisation test (GPMT) of Magnusson & Kligman (24) and the occluded patch test developed by Buehler (25). These are also the methods which are detailed in OECD Test Guideline 406 (26). In the introductions to both of these documents, mention is made of screening procedures, including SAR, the mouse ear swelling test (MEST) and the murine local lymph node assay (LLNA).

Interpreting data generated by using these screening methods is straight forward. If the screening test gives a positive result, the substance is regarded as a skin sensitiser, and may be classified and labelled with the EU risk phrase R43 ("may cause sensitisation by skin contact"). At present, negative results obtained in screening tests must be confirmed by using one of the recognised guinea pig tests. Both of these methods are based on subjective assessments of the elicitation of skin sensitisation reactions following (topical) occluded patch challenge. The results of the guinea pig tests are thus definitive in that positive results which meet/exceed a certain threshold (30% for the GPMT; 15% for the Buehler test) lead to formal classification, but positive results which are lower than this level do not trigger classification (23).

This is important in that it means that alternative tests which are to be used for classification purposes do not necessarily need to be able to detect the weakest sensitisers, only those that are sufficiently active to trigger formal classification. There is a preference amongst the majority of European competent authorities for data for the GPMT to be submitted, since the GPMT is regarded to be more sensitive than the Buehler test. Consequently, the regulatory authorities in Europe often require any use of the Buehler test to be justified scientifically.

Sensitisation Tests Used for Screening/Safety Evaluation/Risk Assessment Purposes

Although the tests which can be used for screening purposes can be the same as those used for regulatory testing, this need not be the case. Screening methods can be tailored individually to meet specific needs that might depend upon the nature of the chemicals to be evaluated and/or their proposed uses. For example, pharmaceuticals may need to be assessed by using a more rigorous protocol than the GPMT (27-29). Often there will be a need to assess several (related) chemicals, so that only those with acceptable skin sensitisation profiles can be developed further, or to obtain data to support clinical testing. The protocols adopted will depend upon the particular industry and the type of chemical to be tested (for example, 30-32). While guinea pig tests are still commonly used for screening purposes, murine tests, such as the MEST (33) and the LLNA (34), have also been developed and are being used.

In current practice, guinea pig sensitisation tests provide an adequate evaluation of skin sensitisation potential on which to base risk assessment and safety procedures (22, 32, 35). Test methods employed for safety evaluation purposes are most commonly based upon dose-response estimation together with a comparative toxicological approach (32). The methods may well be tailored to meet specific needs, but often the main point of comparison is with an existing database (36). Thus, replacement tests of varying sensitivities may all have their roles, but their usefulness will tend to depend heavily upon the availability of an adequate supporting database.

In Vivo Alternatives: Reduction and Refinement of Animal Procedures

In recent years there has been considerable interest in the development of alternative test methods employing mice for assessment of the sensitisation potentials of compounds. Interest in the use of the mouse dates from the development of quantitative methods for the measurement of contact hypersensitivity reactions in this species. The first, and still the most popular, of these methods was the measurement of challenge-induced increases in the ear thickness of previously sensitised mice (the MEST [33]). While other methods have been developed, most notably the incorporation of radio-labelled cells or proteins into the challenge site, the convenience of ear thickness measurements has resulted in widespread use of the MEST in investigative studies. Studies with mice have contributed significantly to our understanding of the immunobiological processes which characterise contact sensitisation.

In its original form, the MEST used a rigorous induction regimen involving the repeated application of test material to tape-stripped abdominal skin, with the treatment site having been prepared previously by the intradermal administration of adjuvant. Control mice were treated with vehicle alone in an identical manner. Seven days following completion of the induction procedure, both test and control animals were treated on the dorsum of one ear with the test material, and on the contralateral ear with vehicle. The concentrations of the test material used for challenge were selected on the basis of results from sighting studies, conducted previously to demonstrate lack of irritant activity. Changes induced in ear thickness were measured one and two days after the challenge. This method has been used to evaluate more than 70 chemicals which, on the basis of guinea pig and/or human studies, were known to possess different skin sensitising potentials. The results of this first study were encouraging, although only modest changes in ear thickness were recorded with some known sensitisers.

A similar method, the mouse ear sensitisation assay, which also incorporates treatment with adjuvant, was described by Descotes (37). In two subsequent investigations, doubts were raised about the sensitivity of the MEST and its ability to identify chemicals with weak or moderate skin sensitising activities (38, 39). However, in at least one of these studies (38), a modification of the original procedure was used. Recently, it has been proposed that the sensitivities of MEST-type assays can be enhanced, and the need for adjuvant obviated, if mice are maintained on a diet enriched in vitamin A (40, 41). In common with the guinea pig methods, the MEST and related assays are based upon the analysis of challenge-induced dermal reactions in previously sensitised animals. One of the advantages over the guinea pig tests is that the MEST uses a quantitative measurement of this response.

A method which seeks to identify chemical allergens as a function of events which characterise the induction phase of skin sensitisation is the LLNA. As described previously, the induction phase of skin sensitisation is characterised by T-lymphocyte activation and proliferation in draining lymph nodes. Vigorous responses also result in substantial increases in the size and cellularity of the draining nodes. It is the measurement of these events which forms the basis of the murine LLNA. In preliminary studies, various parameters were measured following the topical exposure of mice to chemical allergens. It was found that potent chemical allergens caused changes in all of the parameters measured, inducing increased lymph node weight, the appearance of pyroninophilic cells, and vigorous lymph node cell proliferative responses which were augmented in the presence of IL-2. With weaker allergens, however, changes in node weight were modest or virtually undetectable, and it was concluded that lymph node cell proliferation responses represented the most robust and sensitive correlate of skin sensitising activity (42). For this reason, subsequent studies focused upon assessment of the hyperplastic responses induced in draining lymph nodes.

With the purpose of obviating the need for tissue culture, a modified LLNA has been developed, in which proliferative activity is measured in situ following the intravenous injection of treated mice with radio-labelled thymidine (43). It is this assay, in a slightly modified form, which has been the subject of extensive comparisons with the guinea pig tests, the MEST, and the results of human maximisation tests (44-51). Taken together, these analyses have shown the LLNA to be a reliable method for the identification of significant skin sensitising chemicals.

The LLNA has also been evaluated in international interlaboratory trials. The results have shown the method to be robust, in that it gives comparable results in independent laboratories (47, 52). The development and application of the LLNA have been reviewed in detail elsewhere (34, 53). In summary, the assay offers a number of advantages compared with the guinea pig test methods, not least of which are that the LLNA is relatively rapid and cost-effective. The endpoint is quantitative and the assay does not require the use of adjuvant.

These methods are essential components of the screening procedures for new substances. They enable the reduction of animal suffering and of the number of animals required, and they have now been incorporated into testing strategies. As specified in OECD Test Guideline 406 (26), a positive result in the MEST or LLNA can be accepted in place of results from a GPMT or a Buehler test, but a negative result must be confirmed by undertaking one of the recommended guinea pig tests.

Approaches to the Development and Evaluation of Replacement Alternative Methods

Opportunities for developing non-animal methods for skin sensitisation testing are described below.

Structure-activity Relationships and Rule-based Expert Systems

Chemical Basis of Skin Sensitisation
Contact allergy is one of the pathological conditions in which chemistry plays a particularly important role. Chemical reactions and interactions are involved in all of the biological processes which eventually result in the development of delayed hypersensitivity, for example, during penetration of the cutaneous barrier (which is mainly controlled by the physicochemical properties of the allergen), during the formation of the hapten-protein complex (in which chemical bonding is involved), and during recognition between the antigen and the receptors on T-lymphocytes. To try to avoid the inappropriate use of new allergens, our understanding of the features which characterise haptens has to be improved. To promote work in this field within the European Society of Contact Dermatitis, a working party on the chemical bases of allergic contact dermatitis has been formed and has reported its preliminary findings (54).

Interactions of Haptens with Proteins
Many chemical groups have electrophilic properties and are able to react with various nucleophilic groups on proteins to form covalent bonds; this is probably the major mechanism of antigen formation. A very large proportion of biological macromolecules, especially nucleic acids and proteins, contain electron-rich structures (amino, imino, phenolic-OH, and thiol-SH groups). They are, therefore, able to react with electrophilic chemicals, leading to mutagenic (55) and toxic (56) effects, including allergenic effects. Lysine and cysteine are the nucleophilic targets most often affected, but other amino acids containing nucleophilic heteroatoms (for example, histidine and tyrosine), can also react with electrophiles (57).

It is becoming increasingly apparent that each hapten has its own chemical reactivity pattern, even if the effect of this chemical "selectivity" on the processing of hapten-protein conjugates and the selection and/or activation of T-cells is not known. Thus, DNCB, which reacts readily with lysine residues, and alkane methylsulphonates, which react with histidine residues (58), are both strong sensitisers, while DNTB, which reacts with thiol groups, is considered to be a tolerogen. This dichotomy is even more pronounced in the case of methylated analogues of 3-n-alkylcatechols. Compounds methylated at the C-6 position react with amino groups and are strong sensitisers, while compounds methylated at the C-5 position react with thiol groups and are tolerogens (59, 60).

In recent years, mechanisms involving the formation of radicals have figured increasingly in discussions on the mechanism of hapten-protein binding (61). Several studies which indicate that radical reactions are important for haptens containing hydroperoxide groups have been published (62-64). It might be expected that these molecules would react with totally different amino acids, thereby generating different T-epitopes. Nevertheless, not only the formation of covalent bonds results in allergy. For example, with metal salts the formation of a covalent bond is impossible; in this case, a stable coordination complex must be formed between the metal salt and the electron-rich residues of proteins. These coordination complexes are sufficiently stable, and the protein modification sufficiently great, to lead to an allergic response (65).

Metabolism and Prohaptens
The skin is the site of many metabolic processes, which can structurally modify xenobiotics to which it is exposed. These metabolic processes, which are primarily intended for the detoxification and elimination of compounds, can, in certain cases, convert harmless molecules into derivatives with electrophilic, and therefore allergenic, properties. The metabolic processes mainly involve oxidation reactions mediated by extremely powerful hydroxylase enzymes, such as the cytochromes P450, but monoamine oxidases (which convert amines to aldehydes) and peroxidases are also found in the skin. The latter, when activated by the production of hydrogen peroxide during the oxidative stress which can follow dermal exposure to a xenobiotic, can convert electron-rich aromatic derivatives (aminated or hydroxylated) into quinones, which are powerful electrophiles (66, 67). Enzymatic hydrolyses can also occur, for example, with tuliposides A and B, which are found in the bulb of the tulip (Tulipa gesneriana L.), resulting in the release of the actual allergens, tulipalines A and B, respectively (68, 69).

Non-enzymatic processes, such as exposure to atmospheric oxygen or ultraviolet irradiation, can also induce changes in the chemical structures of molecules. Many terpenes spontaneously auto-oxidise in air, producing allergenic derivatives. In the 1950s it was found that the allergenic activity of turpentine was mainly due to the production of hydroperoxides of one of the monoterpenes, (delta)3-carene. This is the case also for another monoterpene, d-limonene, which is found in citrus fruits. d-Limonene itself is not allergenic but, on exposure to air, hydroperoxides, epoxides, and ketones are formed which are strong allergens (70, 71). The diterpenoid resin acids in colophony are also activated by auto-oxidation. The main constituents, abietic acid and dehydroabietic acid, are converted into highly reactive hydroperoxides and epoxides upon contact with air (62, 72-74). These studies show that different haptens can be generated from the same prohapten, and that these do not necessarily cross-react.

Prohaptens play an important role in contact allergy, because of their number and the highly reactive nature of the haptens which they yield. Moreover, as the structures of the metabolites can be very different from the structure of the parent compound, investigations become even more complicated.

Haptens and Cross-allergy
The factors which control molecular recognition during the elicitation stage are primarily the nature of the chemical group and the compatibility of the spatial geometry. Although the nature of the chemical group is very important, and serves to define what are commonly called the "group allergies", it cannot account for all of the SARs. T-cell receptor molecules are highly sensitive to volume and shape, and molecules must have a similar size and spatial geometry to be recognised by the same receptor. The term "cross-allergy" is often misused; its use should be restricted to well-defined cases which are true cross-allergies (75, 76).

True cross-allergy between a sensitiser, A, and a triggering agent, B, can be interpreted in various ways:

  1. A and B are chemically and structurally similar haptens;
  2. A is metabolised to a hapten which is similar to B;
  3. B is metabolised to a hapten which is similar to A; or
  4. A and B are metabolised to similar haptens.

This is important to consider during the development of water-soluble analogues of lipophilic haptens. The chemical modification leading to the water solubility should be sufficiently far from the reactive site of the molecule not to perturb the molecular recognition of the haptenised peptide by T-cell receptors, or both molecules should lead to the same epitope. During a project supported under the European Commission's BRIDGE (Biotechnology Research for Innovation, Development and Growth in Europe) programme (1990-1994), Mabic & Lepoittevin developed water-soluble analogues of pentadecyl catechols based upon two approaches:

  1. the introduction of a water-soluble functional group distant from the reactive site; and
  2. preparation of pro-drugs by the introduction of water-soluble groups which are subsequently removed (by cellular metabolism, due to pH changes, etc.), thereby releasing the original hapten (77).

Some of these analogues have been shown to cross-react with the original hapten (Mabic & J.P. Lepoittevin, unpublished observations).

Structure-activity Relationships for Contact Allergens
The principles underlying the development of QSARs are based on the premise that the properties of a chemical are implicit in its molecular structure. A QSAR is a model which relates the biological activities of a series of similar compounds to one or more physicochemical properties of the compounds (78). In the area of contact allergy, the most important physicochemical features are the reactivity of a potential hapten and its ability to permeate the skin.

Whilst it is true that some limited consideration was given in earlier decades to the relationship between chemical structure and the ability to cause contact allergy, it is only in fairly recent times that a sustained effort has been made in this direction. An important turning point was early in the 1980s, when Dupuis & Benezra (5) described, from an organic chemistry viewpoint, the key features which render a hapten protein reactive. This was closely followed by the description by Roberts & Williams (79) of an approach to the mathematical modelling of skin sensitisation potential, involving the relative alkylation index (RAI). In essence, this model proposed that the degree of sensitisation was closely correlated with the extent to which skin proteins became haptenated. This concept has since been developed and extended. An important area at the present time is the development of new computer-based systems for the identification of skin sensitisers. These are the so-called "expert systems" which use mechanistically based rules to make predictions about the likely effects of novel and existing chemicals.

The relationship between chemical structure and biological activity for skin sensitisation is most apparent for families of organic molecules with similar characteristics, such as the nitrohalobenzenes or the para-substituted benzenes (80). However, of greater importance may be those families of chemicals for which the mechanism of reaction with skin proteins can be postulated to be the same. It is this aspect which was discussed in detail in the original work of Dupuis & Benezra (5). Nevertheless, our understanding of this area of science is far from complete. Apparently straight forward predictions of reaction mechanisms have proven difficult to substantiate in practice (81, 82). Complications arise from the need to understand the role of skin metabolism (83, 84) and the presence of strongly sensitising contaminants (for example, diethyl fumarate in malathion [85], and oxidation products in colophony [74]).

QSAR models of skin sensitisation are based on the premise that the important factors governing the extent/degree of sensitisation are:

  1. the reactivity of the chemical;
  2. its ability to reach the critical epidermal environment; and
  3. the dose applied.

These assumptions form the basis of the RAI model (79), which was used initially to develop equations describing the skin sensitising activities of sultones and p-nitrobenzyl halides (86). It has since been adapted for alkyl alkane sulphonates (87, 88), acrylates (89) and, most recently, a family of furanone derivatives (90, 91).

In general, RAI models use the molar dose and log P (where P is the octanol-water partition coefficient) as two of the parameters. When skin penetration is limited by high hydrophobicity, a (log P)2 term can be valuable (89, 92). However, the chemical reactivity term can be rather difficult to obtain. The best option for obtaining an estimate of reactivity, which was proposed in the original paper (79), is to measure the reactivity of the test material with a defined nucleophile in vitro, and this has been done with a recent series of chemicals (90, 91). Unfortunately, this is not always an easy process, and the choice of acceptor nucleophile may be more important than has been realised previously. Certainly, it is only in a few special cases that we have any knowledge of the nature of the important acceptor sites in vivo (93).

In some cases, QSARs have been derived in which the chemical reactivity was not a statistically significant variable in the relationships. For a series of bromo-alkanes, the reactivity was assumed to be constant (94). For a series of phenyl benzoates, the reactivity, whilst certainly not constant, did not contribute to the model (95). Both of these QSARs rely entirely upon skin permeability parameters. In the former case, the relationship is:

log(1/TD) = 1.61 log P - 0.09 (log P)2 - 7.4
(n = 9; r2=0.94)

where TD is the threshold dose for sensitisation in the LLNA. In the phenyl benzoates example, the relationship is:

TES = 26.48 log P - 0.394 MV - 6.673
(n = 11; r2=0.855)

where TES is the total erythema score in the modified single injection adjuvant test (31) and MV is the molecular volume.

Other groups are known to be applying related approaches, for example neural network analysis of databases of chemicals, grouping chemicals into families with similar chemistries, and searching for common structural elements (96); the majority of these studies are, as yet, unpublished.

Databases
Given a large database, it is possible to carry out a statistical evaluation of the information. This has been done in one case by analysing GPMT results for almost 300 chemicals, and subsequently undertaking multivariate QSAR analysis (97). The outcome was only partially successful, with a discriminant analysis model correctly predicting 88% of non-sensitisers, but having a lower prediction rate for sensitisers (78%).

An extensive database is currently being developed by CCS Associates (Palo Alto, CA; this contains information extracted from articles which have appeared in the journal Contact Dermatitis since 1975. The test results are primarily for single chemicals of known structures, and the database contains chemicals identified by their names, structures, and CAS Registry Numbers, together with information from the tests for contact allergy conducted with them. A structure-activity tree has been developed, based on sets of fragment structures representing the classes and subclasses judged to be relevant for allergic contact dermatitis. These structures are stored in PROPHET, and are used in a computer procedure which is able to assign chemicals to the appropriate structural groups within the structure-activity tree (98, 99). The current problem is to incorporate quantitative aspects into such a database.

Expert Systems
This is an area which is still in its infancy and, in this respect, should be distinguished from computer database-type systems (100). A number of computer-based systems for predicting toxicity from chemical structures have become available in the last few years, for example, TOPKAT, MULTICASE, and DEREK (Deductive Estimation of Risk from Existing Knowledge). TOPKAT and MULTICASE incorporate databases and use statistical methods for making their predictions. In contrast, DEREK is an example of a knowledge-based system ("expert system"); predictions are made on the basis of a series of rules which relate chemical structure to toxicity (101, 102). Rules are predefined by existing knowledge. A collaborative effort in the UK has led to the development of an expert system for skin sensitisation, based on the DEREK architecture (103, 104).

Identification of Structural Alerts
Structural alerts for skin sensitisation were derived initially from 135 chemicals contained in the Unilever historical database (consisting of a total 294 chemicals) which were classified as strong or moderate sensitisers; these were the chemicals which currently would be classified as skin sensitisers according to EU criteria (105). For the identification of structural alerts, the chemicals were divided into groups, on the basis of reaction mechanisms or by empirical derivation:

  1. acylating agents;
  2. alkylating/arylating agents;
  3. "Michael" electrophiles and precursors;
  4. aldehydes and precursors;
  5. free radical generators;
  6. "thiol-exchange" agents; and
  7. others (empirical).

Forty rules (structural alerts) were identified from these groups of chemicals and were programmed into the DEREK system. The DEREK skin sensitisation rulebase has now been extended to contain about 50 rules.

Processing Structures Through DEREK
The user communicates with DEREK by drawing the two-dimensional chemical structure of the query molecule on the screen. The rulebase is then searched against the query structure and any structural alert is highlighted, together with a message indicating the nature of the toxicological hazard (in this case, skin sensitisation). Reference screens are also available which provide literature support for the structural alert which has been triggered.

Of the 135 sensitisers used to generate the original rulebase, DEREK identifies structural alerts likely to lead to skin sensitisation in 133 compounds. The two sensitisers which are not identified by the rulebase as containing structural alerts for skin sensitisation are abietic acid and cinnamic alcohol. The sensitisation potentials of both of these chemicals are believed to be due to oxidation products: 7-oxodehydroabietic acid and 15-hydroxy-7-oxodehydroabietic acid in abietic acid (106), and cinnamic aldehyde in cinnamic alcohol (94). Both of these sensitising impurities trigger structural alerts in the DEREK system. This highlights one difficulty with the approach, namely that sensitisation due to the presence of impurities can only be identified if the chemical structures of those impurities are known.

Of the 120 negative chemicals and the 39 weak sensitisers (chemicals showing some response, but not classifying as sensitisers), the DEREK sensitisation rulebase identified structural alerts in 22 and 16 of the chemicals, respectively. The reason for these apparent "false positives" lies in the fact that skin sensitisation potential depends not just on the ability of the compound to react with a protein either directly or after appropriate biochemical transformation, but also on its ability to penetrate the skin.

Testing the DEREK Rulebase
An assessment was made of the list of 25 positive (that is, sensitisers) and 12 negative substances specifically proposed by the European Centre for Ecotoxicology and Toxicology of Chemicals (ECETOC) for the purposes of novel test evaluation (107). Of the 25 sensitising chemicals, all but one are identified by the DEREK rulebase as containing structural alerts for skin sensitisation (54). The chemical which was not identified by DEREK as containing a structural alert for skin sensitisation, cocoamidopropylbetaine, has recently been shown to be a sensitiser due to the presence of 3-dimethylaminopropylamine (108), a chemical which is used in the manufacture of cocoamidopropylbetaine. Of the 12 chemicals not classified in the EU as skin sensitisers, only two were identified by the DEREK rulebase as containing structural alerts. One of these chemicals, p-aminobenzoic acid, is unlikely to be a significant skin sensitiser due to its poor ability to penetrate the skin. Hydroxypropyl methacrylate met the DEREK criteria for skin sensitisation and was also assessed to be "moderate" for skin penetration, and is thus considered to be a false positive. The remaining ten chemicals did not contain structural alerts for skin sensitisation.

Although still under development, the DEREK skin sensitisation rulebase, together with an assessment of the likely skin permeability of the chemical (109), could be used as a first step in a testing strategy approach for the identification of contact allergens.

Future Trends
It seems likely that, from all of these activities, a consensus will emerge as to the spectrum of common reaction mechanisms and structural alerts for skin sensitisation. The contribution of QSAR and expert systems could be as screens for identifying positive compounds, to assist in the design of new chemical structures, and for the retrospective analysis of large data sets of chemicals. For evaluating the sensitising potentials of new chemicals, these systems could be integrated into an overall testing strategy. However, a number of aspects should be taken into consideration:

  1. the source and quality of the in vivo data used (the chemical purity and knowledge of the impurities present, are important considerations, as well as the need for information supporting the predictive capacity of the in vivo tests used to generate the data); and
  2. the rules already introduced into the expert systems, and any future refinements of these which are incorporated, should be carefully checked and endorsed by an international committee of experts on sensitisation.

The limitations of such systems are their inability to predict effects for entirely new chemistries, the need to take into account skin penetration and metabolism, and the inability to use them for investigating formulations (situations in which there may be antagonism or synergy between several chemicals). Another difficulty is the standardisation/prevalidation of such systems, as is true of all computer-based systems. There is a need to discuss the validation of these systems further. Assessments of them should be made by independent experts (perhaps by ECVAM itself, or by an academic laboratory), and the rules accepted should then be "frozen". If new rules are developed, further validation would be required. In undertaking such a task, all of the systems available should be examined, as well as the different types of software used in their development.

Biological Test Systems

IL-1β Production by Dendritic Cells
IL-1β has been identified as a critical mediator of the induction phase of contact sensitivity (10). IL-1β has been shown to be a product of LC, which is up-regulated within 15 minutes following the epicutaneous application of an allergen to murine skin (9). The effect on IL-1β production was shown to be allergen-specific, since the application of vehicle, (presumed) tolerogen, or irritant, did not result in enhanced IL-1β mRNA expression. These mRNA findings correlated with the analysis of cytokine proteins. In order to causally link epicutaneous sensitisation to the production of IL-1β by epidermal LC, IL-1β was injected into murine skin. Four hours later the resulting epidermal cytokine mRNA signals were analysed by using a quantitative polymerase chain reaction (PCR) procedure, and these were compared with cytokine signals induced by applying hapten epicutaneously to the skin. The injection of IL-1β, but not of control substances (TNF-a, IL-1a or phosphate-buffered saline) resulted in a cytokine pattern which mimicked the cytokine induction which resulted from the application of an allergen epicutaneously to the skin.

To substantiate further the essential role of LC-derived IL-1β in the induction of primary immune responses in skin, epidermal cells (EC) were cultured for two days and all of the non-adherent cells (including all of the LC) were rinsed off the culture plates. Resulting keratinocyte monolayers were then stimulated with water-soluble hapten and cytokines. Afterwards, extracted RNA was analysed for the expression of cytokines by quantitative PCR. Again, only stimulation with IL-1β resulted in a cytokine pattern which mimicked that resulting from the application of an allergen epicutaneously to murine skin, namely the induction of IL-1a, TNF-a, macrophage inflammatory protein (MIP)-2 and IL-10 (10). In marked contrast, stimulation of keratinocyte cultures with a water-soluble hapten did not result in cytokine induction in the absence of LC (that is, in the absence of IL-1β production).

When epidermal sheet and EC suspensions with anti-MHC class II monoclonal antibody (mAb; used for staining LC) were analysed after either injection of IL-1β or epicutaneous application of an allergen, the injection of IL-1β mimicked the changes in LC morphology, density, and MHC class II expression caused by epicutaneous hapten application. Functionally, LC derived from IL-1β-injected skin were more potent accessory cells in an anti-CD3 proliferation assay, using naive CD4+ T-cells as responders, than were LC which had been taken from control skin. In addition, the injection of an anti-IL-1β mAb (but not of control mAb) prior to applying an allergen epicutaneously to murine skin prevented the sensitisation of the mice. Taken together, these data indicate that LC-derived IL-1β is an essential mediator of primary immune responses in skin. The finding that the induction of IL-1β mRNA in LC is allergen-specific provides a useful basis for the development of a DC-based in vitro test system for measuring the allergenic potencies of chemicals.

In Vitro Culture Systems for Dendritic Cells
In the past, the development of LC-based in vitro assays has been limited by the availability of this rare cell type. Recently, in vitro culture systems for DC derived from human peripheral blood have become available (110-112). With regard to phenotype and function, these blood-derived DC share some of the properties of LC derived from human or mouse epidermis. These cells express high levels of MHC class II molecules and certain costimulatory molecules, and do not express macrophage (CD14), B-cell (CD19) or T-cell (CD2) markers. Functionally, they are very potent inducers of primary mixed lymphocyte responses (MLRs). In terms of their cytokine production, they are capable of producing IL-1β (mRNA and protein) following stimulation.

Some laboratories have developed culture systems to reliably generate short-term DC lines from pooled human buffy coats or peripheral blood. Following 7-8 days of culture in the presence of granulocyte-macrophage colony-stimulating factor (GM-CSF) and IL-4, these DC expressed high levels of MHC class II antigens, B7-1, B7-2, CD1a, and CD4O, and were CD2-, CD3-, CD14-, and CD19-. The cell cultures were about 80-90% DC. Functionally, the cells were capable of inducing primary MLRs down to dilutions of 1:640 with respect to the APC:T-cell ratio. When these cells were stimulated with allergens (for example, DNFB, TNCB, TNBS, pentadecylcatechol, and primin) or irritants (for example, sodium dodecyl sulphate, and benzalkonium chloride) at sub-toxic concentrations, and the mRNA levels for β-actin and IL-1β were determined by quantitative reverse-transcriptase (RT)-PCR 15 minutes later, only stimulation with allergens resulted in a significant induction of IL-1β mRNA signals.

These results provide some support for the feasibility of developing an in vitro test system for the detection of allergens based on IL-1β production by DC, although further standardisation of the culture system is necessary (113, 114). A minimal degree of water-solubility of the test material is required with this assay (emulsion testing is possible). The results indicate that this method warrants more widespread evaluation.

As a complementary approach, changes in the endocytosis of MHC class II molecules of blood DC when exposed to contact allergens or irritants could be studied at the same time. It has been shown that murine LC express a diffuse endocytosis pattern for MHC class II molecules when exposed to contact sensitisers, whereas a polar endocytosis pattern is observed following stimulation with irritants (115). The modulation of HLA-DR expression on human blood-derived DC exposed to contact allergens and irritants has also been studied in a short-term incubation system (116). Although only preliminary data are available for human blood DC lines, these suggest that the potential of this approach should be investigated further.

Skin Explant Cultures
Evidence from studies undertaken with a mouse model suggest that LC play the pivotal role in the induction of T-cell mediated immune responses in the early phase of sensitisation (117-119). These epidermal dendritic APC become modified upon capturing hapten. Then, hapten-modified LC migrate from the epidermis to the regional lymph nodes where they stimulate hapten-specific T-cells (120). Thus, the migration of LC from the epidermis via different skin lymph vessels to the lymph nodes is commonly regarded as the critical step in sensitisation.

Studies have shown that the migration event of hapten-modified LC in contact sensitisation occurs in parallel with profound phenotypic changes, for example, the modulation of MHC class II and CD1a+ cell surface molecules (115, 121). In an attempt to elucidate the mechanism of contact sensitisation, studies in animals and with isolated human cells in vitro have demonstrated that, in addition to the primary interactions between LC and CD4+/TCR+ T-cells, the elicitation of delayed hypersensitivity reaction responses requires the involvement of soluble mediators (for example, IL-1β, TNF-a, TGF-β) and several accessory cell-surface adhesion molecules (LFA-1, LFA-3, ICAM-1, VLA-4, E-selectin, and VCAM-1 [9, 10, 17, 21, 122-126]). Penetration of chemicals through the epidermis, and the formation of prohapten, are also important considerations.

The migration process of LC, which could be the critical step for the early phase of contact sensitisation, cannot be studied in classical cell culture systems. Thus, organotypic cultures of human skin explants have been developed for this purpose. Using a reconstituted basement membrane, Kobayashi et al. (127) have developed an elegant assay for assessing the migratory function of LC with respect to allergic contact dermatitis. The use of reconstituted human epidermis in vitro also appears to be promising (128-130).

Recently, a human skin explant culture (HSEC) system (130) has been used to study sensitisation and its regulation by cytokines (131-133). It was demonstrated that the de novo characteristics and distribution of LC in epidermis can be preserved for four days in culture. Early immunological events occuring in the epidermis following the epicutaneous application of known contact allergens, irritants and non-immunogenic chemicals to human skin have been studied (134). The methods used were immunohistochemical morphometry and fluorescence-activated cell sorting (FACS) analysis. After examining seven known allergens, four irritants, and two non-immunogenic compounds, it was shown that epicutaneous application of neither irritants nor tolerogens induced a decrease in the total number of CD1a+/HLA-DR+ epidermal LC in the explants. Only the allergens induced a decrease in the total number of CD1a+/HLA-DR+, which appears to be related both to the migration of LC from the explants and down-regulation of LC-specific HLA-DR and CD1a molecules.

The reduction of CD1a+/HLA-DR+ LC in the epidermis induced by specific allergens appears to parallel the modulation of ICAM-1, VCAM-1, Sialyl Lewis and HECA-452. Interestingly, the early in situ expression of LC-specific IL-1β within four hours after the epicutaneous application of allergens (but not of irritants and tolerogens) appears to coincide with the reduction of epidermal LC (133).

In conclusion, it appears that kinetically monitoring the expression of IL-1β in conjunction with measuring the disappearance of CD1a+/HLA-DR+ LC at early time-points following exposure to test materials, in the ex vivo HSEC model could provide an alternative method for evaluating their sensitisation potentials. In this model, additional parameters, such as changes in neuropeptides, penetration, and the production of metabolites, could also be studied. A prevalidation study of the HSEC system, which is being undertaken in a double-blind manner, is under way. The preliminary results indicate that using the parameter of disappearance of CD1a+/HLA-DR+ LC, the test system was able to distinguish between strong and moderate sensitisers, although false positive and false negative results were obtained with one irritant and one allergen, respectively. It appears that further research is needed to refine the method. The main limitation of this system could be the availability of fresh skin.

Co-culture Systems
Primary contact of T-cells with antigen complexes expressed on APC results in the activation and proliferation of antigen-specific T-lymphocytes. Hauser & Katz (2) first reported the ability of murine LC cultures to sensitise naive autologous T-cells following treatment with TNP or fluorescein isothiocyanate (FITC) in vitro. These results have been reproduced using murine epidermal cells (135) and human epidermal LC (3). Thus, unlike freshly isolated LC, human LC cultured for two days were able to induce proliferation of naive autologous T-lymphocytes in response to TNP and FITC. This proliferative response was specific for the sensitising compound, as demonstrated in a secondary T-cell reaction.

As contact hypersensitivity is a T-cell mediated immune reaction, this in vitro assay may provide a good approach for assessing the sensitising potentials of chemicals. The in vitro assays need to include all of the components which are required for optimal T-cell activation upon presentation of haptens in combination with APC. Comparative studies should be undertaken to optimise several parameters, such as:

  1. culture medium components;
  2. types of APC;
  3. optimal haptenisation;
  4. kinetics of the T-cell response; and
  5. cytokine production.

The water-solubility of the test compound may not be a major limitation with this assay; successfully modified APC and proliferative secondary T-cell responses have been obtained with lipophilic compounds such as DNCB (C. Hauser, unpublished observations), and primary T-cell responses have been observed with DNFB (J.L. Garrigue, unpublished observations).

Most of the data available have been obtained with strong haptens (for example, TNP, FITC, TNCB, DNFB, DNCB) which, in general, humans are not exposed to. Thus, these methods may not be well suited for the determination of the lower allergenic potentials of the majority of contact sensitisers. However, a recent study using human LC as APC was able to discriminate strong and weak sensitisers, and irritants (136).

A limitation of these proliferation assays may be that autologous T-cell responses could hamper the detection of weak contact sensitisers. Furthermore, such assays will not enable the detection of most allergens which are prohaptens and, therefore, require biotransformation in the skin for conversion into the immunogenic hapten. The source of the APC can be a problem with human LC, and thus the development of cell lines should be encouraged.

In conclusion, further research is needed to assess whether T-cell proliferation in vitro represents an assay which will permit allergens and irritants to be distinguished, and which will enable the sensitising potentials of new chemicals to be assessed by a rapid screening test.

Keratinocyte Culture Systems
The in vivo response of skin to external stimuli is a complex process in which various skin cells (endothelial cells, keratinocytes, fibroblasts, melanocytes, LC, Merkel cells, etc.) are involved. Keratinocytes comprise about 95% of the epidermal cells, and they play a key role in skin inflammatory and immunological reactions (137, 138). They act as both primary signal receivers and signal transducers translating exogenous stimuli into endogenous signals (139), and thereby initiating and controlling the body's response at the tissue and systemic levels. Indeed, it has been demonstrated that keratinocytes produce, either constitutively or following induction, numerous proinflammatory mediators (ILs, growth factors, arachidonic acid metabolites, etc.).

The cytokines produced by human keratinocytes are listed in Table I. By virtue of their anatomical location, keratinocytes are the first cells which will come into contact with external stimuli. For this reason, they represent an attractive test system for assessing the cutaneous toxic potentials of chemicals. Several alternative methods which employ keratinocytes have been proposed, in particular for screening chemicals for irritancy (144-146). Keratinocyte cultures have also been proposed as alternative methods for determining the allergenic potentials of chemicals. However, since many cutaneous chemical allergens are also primary irritants, simple in vitro methods may fail to discriminate between the two events.

Table I: Cytokines Expressed by Human Keratinocytes

Cytokine Constitutive expression
Interleukin-1a +
Interleukin-1β +
Interleukin-1 receptor antagonist +
Interleukin-3 +
Interleukin-6 +
Interleukin-8 -a
Interleukin-10 -
Interleukin-12b +
Interferon-induced protein-10 -
Monocyte chemotaxic and activating factor +
Macrophage inflammatory protein-2 +
Colony stimulating factors (GM-CSF, G-CSF) +
Transforming growth factor-a +
Transforming growth factor-β +
Tumour necrosis factor-a -
Interferon-a +
Interferon-Xc -

The information is taken from references 140 and 141.

acytokine expression is not constitutive but can be induced by certain stimuli
bsee reference 142
csee reference 143

GM-CSF = granulocyte-macrophage - colony-stimulating factor; G-CSF = granulocyte - colony-stimulating factor

Two different human keratinocyte culture systems have been used to date:

  1. conventional, submerged (two-dimensional) cultures; and
  2. cultures grown at the air-liquid interface (three-dimensional).

Conventional cell cultures, in which the cells are submerged in culture medium, may be used to study direct interactions between the test material and the living cells (144). The extent of keratinocyte maturation in these cultures is lower than that of the epidermis in vivo, which permits the effects of test materials to be studied on basal and suprabasal cells only. This model is limited to testing water-soluble compounds, and so a large number of preparations with which the skin may come into contact cannot be tested. Furthermore, it is known that the time-course and concentration-dependence of the response in vivo is strongly related to the barrier capacity of the stratum corneum. A good in vitro model must, therefore, adequately mimic the skin barrier function. Hence, test materials should ideally be tested on a cultured skin substitute which is morphologically, biochemically, and functionally representative of its in vivo counterpart. This condition can be met to a large extent when keratinocytes are cultured at the air-liquid interface, attached to an appropriate substrate (147, 148).

When keratinocytes are cultured at the air-liquid interface, a fully differentiated epidermis with a coherent horny layer is formed. The presence of the stratum corneum and exposure to air permit a realistic, topical application of various compounds on the reconstructed epidermal surface. In this way, the effects of skin irritants have been evaluated by assessing changes in:

  1. tissue morphology;
  2. cell viability;
  3. keratin expression;
  4. lipid composition;
  5. barrier function;
  6. expression of cell adhesion molecules (145); and
  7. cytokine production (149).

Keratinocytes have been shown to produce cytokines which are involved in the induction of skin sensitisation. At present, it is not clear whether it is possible with this in vitro system to discriminate between skin sensitisers and irritants. Studies directed toward answering this question should examine the kinetics and dose-dependence of the induction of appropriate mediators following the application of allergens and irritants. Since interaction between keratinocytes and LC in vivo may determine the final biological effect, a model in which keratinocytes are co-cultured with DC should be developed and evaluated for its potential to discriminate between skin sensitisers and irritants, in order to establish a test system specifically for the detection of allergens.

In conclusion, progress has been made in understanding the immunological mechanisms underlying skin sensitisation. However, the most critical aspect is to identify relevant endpoints for predicting the sensitising potentials of new chemicals. The markers should be specific to the sensitisation process (that is, they should not respond in the presence of irritants), and they should be sufficiently sensitive to respond to moderate and mild sensitisers, not only to strong sensitisers.

Much of our current knowledge derives from experiments conducted with mice. The possibility to use human cells/tissues means that we should be in a better position to understand the actual sensitisation mechanisms occuring in humans. However, this may concomitantly introduce a difficulty with respect to the standardisation of the alternative tests, due to the great heterogeneity of the human response. This problem could possibly be addressed by the development of relevant cell lines of human origin.

Conclusions

Progress has been made in terms of reducing animal numbers and refining those animal procedures which are undertaken (for example, 33, 34, 50, 52, 150-153). However, animal experiments are still necessary to assess the skin sensitising potentials of new chemicals. At this time, only in vivo tests provide the information needed for conducting safety evaluations and risk assessments.

Advances have been made in our basic understanding of skin sensitisation mechanisms. Some specific areas, such as the application of QSAR and cell culture systems, clearly show promise. However, in vitro culture methods which could be developed at this stage will suffer from solubility problems, metabolic deficiencies, and the lack of appropriate barrier properties. Thus, it is highly likely that any overall prediction of skin sensitisation potential will only be achieved by combining information obtained in several, complementary tests.

The way forward could be the standardisation of some of the most promising assays, such as the production of IL-1β by human DC, followed by their prevalidation. The next step would then be the development of a sufficient database with these assays, via a validation exercise against an appropriate in vivo database. This could include the human database generated by Kligman (154), which contains results for 80 chemicals tested in the human maximisation test. The databases on the GPMT (97, 155) represent another source of in vivo data. While there are several promising non-animal alternative methods, there remains a need for basic research to be conducted alongside any prevalidation studies.

Recommendations

Structure-Activity Relationships and Rule-based Expert Systems

  1. The current expert system rulebase in DEREK should be used as a screen for skin sensitisation potential.
  2. The existing rulebase for skin sensitisation should be peer reviewed by an independent group of experts, in order to gain wider acceptance of recommendation 1 above.
  3. It is recognised that the DEREK sensitisation rulebase will be further refined in the future. These developments must be peer-reviewed and endorsed before implementation.
  4. The development of additional systems based on SARs for the detection of skin sensitisers should be encouraged.

Biological Test Systems

  1. The induction of IL-1β expression in cultures of human dendritic cells as a screen for potential skin sensitisers should be evaluated. Methods for the isolation and culture of human peripheral blood dendritic cells must be standardised. This should also enable other chemical-induced changes, which could be of value in the assessment of sensitising capacity, to be investigated.
  2. There is a need to develop and evaluate other models for use in the assessment of sensitising activity. An explant method shows promise and is being evaluated currently. Decisions regarding the applicability of this method should be made once the results of this study are available.
  3. Methods need to be developed which enable the introduction of lipophilic chemicals into culture medium.
  4. A skin metabolising system should be developed for use in concert with in vitro culture methods.
  5. Basic research in chemical allergy and skin sensitisation should be encouraged, and additional funding must be made available. Funding should also be made available for the development, evaluation, and exploitation of new test methods.

References

  1. Anon. (1994). ECVAM News and Views. ATLA 22: 7-11.
  2. Hauser, C. & Katz, S.I. (1988). Activation and expansion of hapten- and protein-specific T-helper cells from non-sensitized mice. Proceedings of the National Academy of Sciences USA 85: 5625-5628.
  3. Moulon, C., Peguet-Navarro, J., Courtellemont, P., Redziniak, G. & Schmitt, D. (1993). In vitro primary sensitization and restimulation of hapten-specific T-cells by fresh and cultured human epidermal Langerhans cells. Immunology 80: 373-379.
  4. Hauser, C. (1990). Cultured epidermal Langerhans cells activate effector cells for contact sensitivity. Journal of Investigative Dermatology 95: 436-440.
  5. Dupuis, G. & Benezra, C. (1982). Allergic Contact Dermatitis to Simple Chemicals. pp. 66-68. New York: Marcel Dekker, Inc.
  6. Martin, S. & Weltzien, H.U. (1994). T-cell recognition of haptens, a molecular view. International Archives of Allergy and Immunology 104: 10-16.
  7. Weltzien, H.U., Moulon, C., Martin, S., Padovan, E., Hartmann, U. & Kohler, J. (1996). T-cell immune responses to haptens: structural models for allergic and autoimmune reactions. Toxicology 107: 141-151.
  8. Cavani, A., Hackett, C.J., Wilson, K.J., Rothbart, J.B. & Katz, S.I. (1995). Characterization of epitopes recognized by hapten-specific CD4+ T-cells. Journal of Immunology 154: 1232-1238.
  9. Enk, A.H. & Katz, S.I. (1992). Early molecular events in the induction phase of contact sensitivity. Proceedings of the National Academy of Sciences (USA) 89: 1398-1402.
  10. Enk, A.H., Angeloni, V.L., Udey, M.C. & Katz, S.I. (1993). An essential role for Langerhans cell-derived IL-1β in the initiation of primary immune responses in skin. Journal of Immunology 150: 3698-3704.
  11. Cumberbatch, M. & Kimber, I. (1995). Tumour necrosis factor-a is required for accumulation of dendritic cells in draining lymph nodes and for optimal contact sensitisation. Immunology 84: 31-35.
  12. Roake, J.A., Raos, A.S., Morris, P.J., Larsen, C.P., Hankins, D.F. & Austyn, J.M. (1995). Dendritic cell loss from non-lymphoid tissues after systemic administration of lipopolysaccharide tumor necrosis factor and interleukin 1. Journal of Experimental Medicine 181: 2237-2247.
  13. Schuler, G. & Steinman, R.M. (1985). Murine epidermal Langerhans cells mature into potent immunostimulatory dendritic cells in vitro. Journal of Experimental Medicine 161: 526-546.
  14. Aiba, S. & Katz, S.I. (1992). Phenotypic and functional characteristics of in vivo activated Langerhans cells. Journal of Immunology 145: 2791-2796.
  15. Griffiths, C., Barker, J. & Nickoloff, B. (1991). Modulation of leukocyte adhesion molecules ELAM-1, ICAM-1, VCAM-1, T-cell chemotaxin, IL-8, and TNF-a in allergic contact dermatitis. Clinical Research 39: 194A.
  16. Bucksbaum, M.E., Kupper, T.S. & Murphy, G.F. (1993). Differential induction of intracellular adhesion molecule-A in human skin by recombinant cytokines. Journal of Cutaneous Pathology 20: 21-27.
  17. Silber, A., Newman, W., Sasseville, V.G., Pauley, D., Beall, D., Walsh, D.G. & Ringler, D.J. (1994). Recruitment of lymphocytes during cutaneous delayed hypersensitivity in non-human primates is dependent on E-selectin and vascular adhesion molecule. Journal of Clinical Investigations 93: 1554-1563.
  18. Goebeler, M., Meinardus, H.G., Roth, J., Goerdt, S. & Sorg, C. (1993). Nickel chloride and cobalt chloride, two common contact sensitisers directly induce expression of intracellular adhesion molecule-1 (ICAM-1), vascular adhesion molecule-1 (VCAM-1), and endothelial cell adhesion molecule (ELAM-1) by endothelial cells. Journal of Investigative Dermatology 100: 759-765.
  19. Askenase, P.W. & Ptak, W. (1992). Role of ysigma-T-cells as positive regulators of aβ-TCR positive effector T-cells in allergic contact sensitivity. International Archives of Allergy and Immunology 99: 246-251.
  20. Gocinski, B.L. & Tigelaar, R.E. (1990). Roles of CD4+ and CD8+ T-cells in murine contact sensitivity revealed by in vivo monoclonal antibody depletion. Journal of Immunology 144: 4121-4128.
  21. Mosman, T.R. (1992). T-lymphocyte subsets, cytokines and effector functions. Annals of the New York Academy of Sciences 664: 89-92.
  22. Botham, P.A., Basketter, D.A., Maurer, T., Mueller, D., Potokar, M. & Bontinck, W.J. (1991). Skin sensitisation - a critical review of predictive test methods in animals and man. Food and Chemical Toxicology 29: 275-286.
  23. Anon. (1992). Annex to Commission Directive 92/69/EEC of 31 July 1992 adapting to technical progress for the seventeenth time Council Directive 67/548/EEC on the approximation of laws, regulations, and administrative provisions relating to the classification, packaging, and labelling of dangerous substances. Official Journal of the European Communities L383A: 131-136.
  24. Magnusson, B. & Kligman, A.M. (1970). Allergic Contact Dermatitis in the Guinea Pig, 141 pp. Springfield: Illinois: Charles C. Thomas.
  25. Buehler, E.V. (1965). Delayed contact hypersensitivity in the guinea pig. Archives of Dermatology 91: 171-177.
  26. Anon. (1993). OECD Guideline for Testing of Chemicals, No. 406. Skin Sensitisation, 9 pp. Paris: OECD.
  27. Maurer, T., Thomann, P., Weirich, E.G., & Hess, R. (1979). Predictive evaluation in animals of the contact allergenic potential of medically important substances. Part II. Contact Dermatitis 5: 1-10.
  28. Maurer, T. (1985). The optimization test. Current Problems in Dermatology 14: 114-151.
  29. Scheper, R.J., von Blomberg, M.B.E., de Groot, J., Goeptar, A.R., Lang, M., Oostendorp, R.A.J., Bruynzeel, D.P. & van Tol, R.G.L. (1990). Low allergenicity of clonidine impedes studies of sensitization mechanism in guinea pig models. Contact Dermatitis 23: 81-89.
  30. Sharp, D.W. (1981). The sensitization potential of some perfume ingredients tested using a modified Draize procedure. Toxicology 9: 261-271.
  31. Goodwin, B.J.F. & Johnson, A.W. (1985). Single injection adjuvant test. In Contact Allergy Predictive Tests in Guinea Pigs (ed. K.E. Andersen & H.I. Maibach), pp. 201-207. Basel, Switzerland: Karger.
  32. Gerberick, G.F., Robinson, M.K. & Stotts, J. (1994). An approach to allergic contact sensitisation risk assessment of new chemicals and product ingredients. American Journal of Contact Dermatitis 4: 205-211.
  33. Gad, S.C., Dunn, B.J., Dobbs, D.W., Reilly, C. & Walsh, R.D. (1986). Development and validation of an alternative dermal sensitization test: the mouse ear swelling test. Toxicology and Applied Pharmacology 84: 93-114.
  34. Kimber, I., Dearman, R.J., Scholes, E.W. & Basketter, D.A. (1994). The local lymph node assay: developments and applications. Toxicology 93: 13-31.
  35. Basketter, D.A. (1994). Guinea pig predictive tests for contact hypersensitivity. In Immunotoxicology and Immunopharmacology (ed. J.H. Dean, M.I. Luster, A.E. Munson & I. Kimber), pp. 693-702. New York: Raven Press.
  36. Basketter, D.A., Gerberick, G.F. & Robinson, M.K. (1996). Risk assessment. In The Toxicology of Contact Sensitivity (ed. I. Kimber & T. Maurer), pp. 152-164. London: Taylor & Francis.
  37. Descotes, J. (1988). Identification of contact allergens: the mouse ear sensitization assay. Journal of Toxicology - Cutaneous and Ocular Toxicology 7: 263-272.
  38. Cornacoff, J.B., House, R.V. & Dean, J.H. (1988). Comparison of a radioisotopic incorporation method and the mouse ear swelling test (MEST) for contact sensitivity to weak sensitizers. Fundamental and Applied Toxicology 10: 40-44.
  39. Dunn, B.J., Rusch, G.M., Siglin, J.C. & Blaszcak, D.L. (1990). Variability of a mouse ear swelling test (MEST) in predicting weak and moderate contact sensitization. Fundamental and Applied Toxicology 15: 242-248.
  40. Thorne, P.S., Hawk, C., Kaliszewski, S.D. & Guiney, P.D. (1994). The non-invasive mouse ear swelling assay. I. Refinements for detecting weak sensitizers. Fundamental and Applied Toxicology 17: 790-806.
  41. Sailstad, D.M., Tepper, J.S., Doerfler, D.L. & Selgrade, M.K. (1993). Evaluation of several variations of the mouse ear swelling test (MEST) for detection of weak and moderate contact sensitizers. Toxicology Methods 3: 169-182.
  42. Kimber, I. (1989). Aspects of the immune response to contact allergens: opportunities for the development and modification of predictive test methods. Food and Chemical Toxicology 27: 755-762.
  43. Kimber, I., Hilton, J. & Weisenberger, C. (1989). The murine local lymph node assay for identification of contact allergens: a preliminary evaluation of in situ measurement of lymphocyte proliferation. Contact Dermatitis 21: 215-220.
  44. Kimber, I., Hilton, J. & Botham, P.A. (1990). Identification of contact allergens using the murine local lymph node assay: comparisons with the Buehler occluded patch test in guinea pigs. Journal of Applied Toxicology 10: 173-180.
  45. Kimber, I., Hilton, J., Botham, P.A., Basketter, D.A., Scholes, E.W., Miller, K., Robbins, M.C., Harrison, P.T.C., Gray, T.J.B. & Waite, S.J. (1991). The murine local lymph node assay: results of an interlaboratory trial. Toxicology Letters 55: 203-213.
  46. Kimber, I. & Weisenberger, C. (1989). A murine local lymph node assay for the identification of contact allergens. Assay development and results of an initial validation study. Archives of Toxicology 63: 274-282.
  47. Kimber, I. & Basketter, D.A. (1992). The murine local lymph node assay: a commentary on collaborative studies and new directions. Food and Chemical Toxicology 30: 165-169.
  48. Kimber, I. & Dearman, R.J. (1994). Assessment of contact and respiratory hypersensitivity in mice. In Immunotoxicology and Immunopharmacology (ed. J.H. Dean, M.I. Luster, A.E. Munson & I. Kimber), pp. 721-732. New York: Raven Press.
  49. Basketter, D.A., Scholes, E.W., Kimber, I., Botham, P.A., Hilton, J., Miller, K., Robbins, M.C., Harrison, P.T.C. & Waite, S.J. (1991). Interlaboratory evaluation of the local lymph node assay with 25 chemicals and comparison with guinea pig test data. Toxicology Methods 1: 30-43.
  50. Basketter, D.A,, Selbie, E., Scholes, E.W., Lees, D., Kimber, I. & Botham, P.A. (1993). Results with OECD recommended positive control sensitizers in the maximization, Buehler, and local lymph node assays. Food and Chemical Toxicology 31: 63-67.
  51. Basketter, D.A., Scholes, E.W. & Kimber, I. (1994). The performance of the local lymph node assay with chemicals identified as contact allergens in the human maximization test. Food and Chemical Toxicology 32: 543-547.
  52. Kimber, I., Hilton, J., Dearman, R.J., Gerberick, G.F., Ryan, C.A., Basketter, D.A., Scholes, E.W., Ladics, G.S., Loveless, S.E., House, R.V. & Guy, A. (1995). An international evaluation of the murine local lymph node assay and comparison of modified procedures. Toxicology 103: 63-73.
  53. Kimber, I. (1993). The murine local lymph node assay: principles and practice. American Journal of Contact Dermatitis 4: 42-44.
  54. Basketter, D., Dooms-Goossens, A., Karlberg, A.T. & Lepoittevin, J.P. (1995). The chemistry of contact allergy: why is a molecule allergenic? Contact Dermatitis 32: 65-73.
  55. Frierson, M.R., Klopman, G. & Rosenkranz, H.S. (1985). Structure activity relationships among mutagens and carcinogens: a review. Environmental Mutagenesis 8: 283-327.
  56. Guengerich, F.P. & Liebler, D.C. (1985). Enzymatic activation of chemicals. CRC Critical Reviews in Toxicology 14: 259-307.
  57. Means, G.E. & Feeney, R.E. (1971). Chemical Modification of Proteins. 254 pp. San Francisco, CA: Holden-day Inc.
  58. Lepoittevin, J.P. & Benezra, C. (1992). 13-C enriched methyl alkanesulfonates: new lipophilic methylating agents for the identification of nucleophilic amino acids of proteins by NMR. Tetrahedron Letters 33: 3875-3878.
  59. Stampf, J.L., Benezra, C., Byers, V. & Castagnoli, N. (1986). Induction of tolerance to poison ivy in the guinea pig by epicutaneous application of the structural analog 5-methyl-3-n-pentadecyl catechol. Journal of Investigative Dermatology 5: 535-538.
  60. Nazih, A., Benezra, C. & Lepoittevin, J.P. (1993). Bihaptens with 5- and 6-methyl-substituted alkylcatechols and methylene lactone functional groups: tools for hapten (allergen or tolerogen) protein interaction studies. Chemical Research in Toxicology 6: 215-222.
  61. Schmidt, R., Kahn, L. & Chung, L.Y. (1990). Are free radicals and quinones the haptenic species derived from urushiols and other contact allergenic mono- and dihydric alkylbenzenes? The significance of NADH, glutathione, and redox cycling in the skin. Archives of Dermatological Research 282: 56-64.
  62. Gäfvert, E., Shao, L.P., Karlberg, A-T, Nilsson, U. & Nilsson, J.L.G. (1994). Contact allergy to resin acid hydroperoxides. Hapten binding via free radicals and epoxides. Chemical Research in Toxicology 7: 260-266.
  63. Lepoittevin, J.P. & Karlberg, A-T. (1994). Interaction of allergenic hydroperoxides with proteins: a radical mechanism. Chemical Research in Toxicology 7: 130-133.
  64. Lepoittevin, J.P., Benezra, C., Sigman, C.C., Bagheri, D., Fraginals, R. & Maibach, H.I. (1994). Molecular aspects of allergic contact dermatitis. In Text Book of Contact Dermatitis(Ed. R.J.G. Rycroft, T. Menné, P.J. Frosch), pp. 105-119. Heidelberg, Germany: Springer-Verlag.
  65. Sinigaglia, F. (1994). The molecular basis of metal recognition by T-cells. Journal of Investigative Dermatology 102: 398-401.
  66. Liberato, D.J., Byers, V.S., Dennick, R.G. & Castagnoli, N. Jr. (1981). Regiospecific attack of nitrogen and sulfur nucleophiles on quinones derived from poison oak/ivy catechols (urushiols) and analogues as models for urushiol-protein conjugate formation. Journal of Medicinal Chemistry 24: 28-33.
  67. Lepoittevin, J.P. & Benezra, C. (1991). Allergic contact dermatitis caused by naturally occuring quinones. Pharmaceutica Weekblad (Sci) 13: 119-122.
  68. Tschesche, R., Kammerer, F.J. & Wulff, G. (1969). Uber die struktur der antiobiotisch aktiven Substanzen der Tulpe (Tulipa gesneriana L.). Chemische Berichte 102: 2057-2071.
  69. Verspyck Mijnssen, G.A.W. (1969). Pathogenesis and causative agents of "tulip fingers". British Journal of Dermatology 81: 737-745.
  70. Karlberg, A-T, Magnusson, K. & Nilsson, U. (1992). Air oxidation of d-limonene - the citrus solvent - creates potent allergens. Contact Dermatitis 26: 332-340.
  71. Karlberg, A-T, Shao, L.P., Nilsson, U., Gäfvert, E. & Nilsson, J.L.G. (1994). Hydroperoxides in oxidized d-limonene identified as potent contact allergens. Archives of Dermatological Research 286: 97-103.
  72. Karlberg, A-T. (1988). Contact allergy to colophony. Chemical identification of allergens. Sensitization experiments and clinical experiments. Acta Dermatologia-Venereologia 68 (Suppl. 139), 1-43.
  73. Karlberg, A-T, Bohlinder, K., Boman, A., Hacksell, U., Hermansson, J., Jacobsson, S. & Nilsson, J.L.G. (1988). Identification of 15-hydroperoxyabietic acid as a contact allergen in Portuguese colophony. Journal of Pharmacy & Pharmacology 40: 42-47.
  74. Karlberg, A-T, Boman, A., Hacksell, U., Jacobsson, S. & Nilsson, J.L.G. (1988). Contact allergy to dehydroabietic acid derivatives isolated from Portuguese colophony. Contact Dermatitis 19: 166-174.
  75. Baer, R.L. (1954). Cross-sensitization phenomena. In Modern Trends in Dermatology (ed. R.M.B. McKenna), pp. 232-252. London: Butterworths.
  76. Benezra, C. & Maibach, H. (1984). True cross-sensitization, false cross-sensitization, and otherwise. Contact Dermatitis 11: 65-69.
  77. Barbier, A., Rizova, E., Stampf, J.L., Lacheretz, F., Pistoor, F.H.M., Bos, J.D., Kapsenberg, M.L., Becker, D., Mohamadzadeh, M., Knop, J., Mabic, S. & Lepoittevin, J-P. (1994). Development of a predictive in vitro test for detection of sensitizing compounds (European BRIDGE project). In In Vitro Skin Toxicology: Irritation, Phototoxicity, Sensitisation (ed. A. Rougier, A.M. Goldberg & H.I. Maibach), Alternative Methods in Toxicology, Volume 10, pp. 341-350. New York: Mary Ann Liebert.
  78. Karcher, W., Hansen, B.G., van Leeuwen, C., Wagner, P. & Auer, C. (1995). Assessment of QSARs for predicting fate and effects of chemicals in the environment: an international European project. SAR and QSAR Environmental Research 3: 217-221.
  79. Roberts, D.W. & Williams, D.L. (1982). The derivation of quantitative correlations between skin sensitization and physicochemical parameters for alkylating agents, and their application to experimental data for sultones. Journal of Theoretical Biology 99: 807-825.
  80. Mayer, R.L. (1954). Group sensitization to compounds of quinone structure and its biochemical basis. Roles of these substances in cancer. Progress in Allergy 4: 79-172.
  81. Basketter, D.A. & Goodwin, B.F.J. (1988). Investigation of the prohapten concept: cross reactions between 1,4-substituted benzene derivatives in the guinea pig. Contact Dermatitis 19: 248-253.
  82. Basketter, D.A. & Lidén, C. (1992). Further investigation of the prohapten concept: reactions to benzene derivatives in man. Contact Dermatitis 27: 90-97.
  83. Martin, R.J., Denyer, S.P. & Hadgraft, J. (1987). Skin metabolism of topically applied compounds. International Journal of Pharmaceutics 39: 23-32.
  84. Scholes, E.W., Pendlington, R.U., Sharma, R.K. & Basketter, D.A. (1994). Skin metabolism of contact allergens. Toxicology in Vitro 8: 551-553.
  85. Hjorth, N. & Wilkinson, D.S. (1968). Contact dermatitis. Sensitization to pesticides. British Journal of Dermatology 80: 272-274.
  86. Roberts, D.W., Goodwin, B.F.J., Williams, D.L., Jones, K., Johnson, A.W. & Alderson, J.C.E. (1983). Correlations between skin sensitisation potential and chemical reactivity for p-nitrobenzyl compounds. Food and Chemical Toxicology 21: 811-813.
  87. Roberts, D.W. & Basketter, D.A. (1990). A quantitative structure activity/dose response relationship for contact allergenic potential of alkyl group transfer agents. Contact Dermatitis 23: 331-335.
  88. Fraginals, R., Roberts, D.W., Lepoittevin, J.P & Benezra, C. (1991). Refinement of the relative alkylation index (RAI) model for skin sensitization and application to mouse and guinea pig test data for alkylsulfonates. Archives of Dermatological Research 283: 387-394.
  89. Roberts, D.W. (1987). Structure activity relationships for skin sensitization potential of diacrylates and dimethylacrylates. Contact Dermatitis 17: 281-289.
  90. Franot, C., Roberts, D.W., Smith, R.G., Basketter, D.A., Benezra, C. & Lepoittevin, J.P. (1994). Structure activity relationships for contact allergenic potential of g,g-dimethyl-g-butyrolactone derivatives. I. Synthesis and electrophilic reactivity studies of a-(w-substituted-alkyl)-g,g-dimethyl-g-butyro-lactones and correlation of skin sensitization potential and cross-sensitization patterns with structure. Chemical Research in Toxicology 7: 297-306.
  91. Franot, C., Roberts, D.W., Basketter, D.A., Benezra, C. & Lepoittevin, J.P. (1994). Structure activity relationships for contact allergenic potential of g,g-dimethyl-g-butyrolactone derivatives. II. Quantitative structure-skin sensitization relationships for a-(w-substituted-alkyl)-g,g-dimethyl-g-butyro-lactones. Chemical Research in Toxicology 7: 307-312.
  92. Basketter, D.A. & Roberts, D.W. (1990). Structure/activity relationships in contact allergy. International Journal of Cosmetic Science 12: 81-90.
  93. Franot, C., Benezra, C. & Lepoittevin, J-P. (1993). Synthesis and interaction studies of 13C-labelled lactone derivatives with a model protein using 13C NMR. Bioorganic and Medicinal Chemistry 1: 389-397.
  94. Basketter, D.A., Roberts, D.W., Cronin, M. & Scholes, E.W. (1992). The value of the local lymph node assay in quantitative structure-activity investigations. Contact Dermatitis 27: 137-142.
  95. Barratt, M.D., Basketter, D.A. & Roberts, D.W. (1994). Skin sensitisation structure activity relationships for phenyl benzoates. Toxicology in Vitro 8: 823-826.
  96. Magee, P.S., Hostynek, J.J. & Maibach, H.I. (1994). A classification model for allergenic contact dermatitis. Quantitative Structure Activity Relationships 13: 22-33.
  97. Cronin, M.T.D. & Basketter, D.A. (1994). Multivariate QSAR analysis of a skin sensitization database. SAR and QSAR in Environmental Research 2: 159-179.
  98. Benezra, C., Sigman, C.C., Perry, B.A., Helmes, C.T. & Maibach, H.I. (1985). A systematic search for structure-activity relationships of skin contact sensitizers: methodology. Journal of Investigative Dermatology 85: 351-356.
  99. Benezra, C., Sigman, C.C., Bagheri, D.C., Helmes, C.T. & Maibach, H.I. (1989). A systematic search for structure-activity relationships of skin contact sensitizers: II. para-Phenylenediamines. Seminars in Dermatology 8: 88-93.
  100. Dooms-Goossens, A., Drieghe, J., Degreef, H. & Dooms, M. (1990). The "Codex-E": an expert system for contact dermatitis. Contact Dermatitis 22: 180-181.
  101. Sanderson, D.M. & Earnshaw, C.G. (1991). Computer prediction of possible toxic action from chemical structure, the DEREK system. Human and Experimental Toxicology 10: 261-273.
  102. Ridings, J.E., Barratt, M.D., Cary, R., Earnshaw, C.G., Eggington, C.E., Ellis, M.K., Judson, P.N., Langowski, J.J., Marchant, C.A., Payne, M.P., Watson, W.P. & Yih, T.D. (1996). Computer prediction of possible toxic action from chemical structure, an update on the DEREK system. Toxicology 106: 267-279.
  103. Barratt, M.D., Basketter, D.A., Chamberlain, M., Payne, M.P., Admans, G.D. & Langowski, J. (1994). Development of an expert system rulebase for identifying contact allergens. Toxicology in Vitro 8: 837-839.
  104. Barratt, M.D. & Basketter, D.A. (1994). Structure activity relationships for skin sensitization: an expert system. In In Vitro Skin Toxicology. Irritation, Phototoxicity, Sensitisation (ed. A. Rougier, A.M. Goldberg & H.I. Maibach), Alternative Methods in Toxicology, Volume 10, pp. 293-301. New York: Mary Ann Liebert.
  105. Anon. (1993). Annex IV to Commission Directive 93/21/EEC of 27 April 1993 adapting to technical progress for the 18th time Council Directive 67/548/EEC on the approximation of the laws, regulations and administrative provisions relating to the classification, packaging and labelling of dangerous substances. Official Journal of the European Communities L11OA: 59.
  106. Karlberg, A-T. (1991). Air oxidation increases the allergic potential of Tall Oil Rosin. American Journal of Contact Dermatitis 2: 43-49.
  107. Kimber, I., Basketter, D.A., Briatico-Vangosa, G., Cookman, G.R., Evans, P., Loveless, S.E. & Pauluhn, J. Skin and Respiratory Sensitizers: Reference Chemicals Data Bank. ECETOC Monograph. Brussels: European Centre for Ecotoxicology and Toxicology of Chemicals. In press.
  108. Angelini, G., Foti, C., Rigano, L. & Vena, G.A. (1995). 3-Dimethylaminopropylamine: a key substance in contact allergy to cocamidopropylbetaine? Contact Dermatitis 32: 96-99.
  109. Barratt, M.D. (1995). Quantitative structure activity relationships for skin permeability. Toxicology in Vitro 9: 27-37.
  110. Romani, N., Gruner, S., Brang, D., Kamgen, E., Lenz, A., Trockenbacher, B., Konwalinka, G., Fritsch, P.O., Steinman, R.M. & Schuler, G. (1994). Proliferating dendritic cell progenitors in human blood. Journal of Experimental Medicine 180: 83-93.
  111. Sallust, F. & Lanzavecchia, A. (1994). Efficient presentation of soluble antigen by cultured human dendritic cells is maintained by granulocyte/macrophage colony stimulating factor plus interleukin-4 and down-regulated by tumour necrosis factor a. Journal of Experimental Medicine 179: 1109-1118.
  112. Lenz, A., Heine, M., Schuler, G. & Romani, N. (1993). Human and murine dendritic cells. Isolation by means of a novel method and phenotypical and functional characterisations. Journal of Clinical Investigations 95: 2587-2596.
  113. Hoppe, U., Degwert, J. & Steckel, F. (1995). Use of CD1a- dendritic cells and keratinocytes to characterize cellular reactions involved in allergic contact dermatitis. Journal of Cellular Biochemistry 21 (Suppl. A), 11.
  114. Jonuleit, H., Müller, G., Degwert, J., Hoppe, U., Knop, J. & Enk, A. IL-1β upregulation in dendritic cells by contact sensitizers: generation of an in vitro test system for contact allergens. XXIII. Jahrestagurg der Arbeits-gemeinschaft Dermatologische Forschung, in press.
  115. Becker, D., Neiss, U., Neiss, S., Reske, K. & Knop, J. (1991). Contact allergens modulate the expression of MHC class II molecules on murine epidermal Langerhans cells by endocytotic mechanism. Journal of Investigative Dermatology 98: 700-705.
  116. Degwert, J., Steckel, F. & Hoppe, U. (1995). Stimulatory capacities of human blood-derived dendritic cells and their phenotypical alterations in the presence of contact sensitizers. Journal of Cellular Biochemistry 21 (Suppl. A), 10.
  117. Streilein, J.W. & Bergstresser, P.R. (1981). Langerhans cell function dictates induction of contact hypersensitivity or unresponsiveness to DFNB in Syrian hamsters. Journal of Investigative Dermatology 77: 272-274.
  118. Willis, C.M., Stephens, C.J. & Wilkinson, J.D. (1990). Differential effects of structurally unrelated chemical irritants on the density and morphology of epidermal CD1+ cells. Journal of Investigative Dermatology 95: 711-716.
  119. Kapsenberg, M.L., Bos, J.D. & Wierenga, E.A. (1992). T-cells in allergic contact responses to haptens and proteins. Springer Seminars in Immunopathology 13: 303-314.
  120. Kripke, M.L., Munn, C.G., Jeevan, A., Tang, J.M. & Bucana, C. (1990). Evidence that cutaneous antigen-presenting cells migrate to regional lymph nodes during contact sensitisation. Journal of Immunology 145: 2833-2838.
  121. van Wilsen, E.J.G., Breve, J., Kleijmeer, M. & Kraal, G. (1994). Antigen bearing Langerhans cells in skin draining lymph nodes: phenotype and kinetics of migration. Journal of Investigative Dermatology 103: 217-220.
  122. Piguet, P.F., Grau, G.E., Hauser, C. & Vassali, P. (1991). Tumour necrosis factor is a critical mediator in hapten induced irritant and contact sensitivity reactions. Journal of Experimental Medicine 173: 673-679.
  123. Scheynius, A., Camp, R.L. & Pure, E. (1993). Reduced contact sensitivity reactions in mice treated with monoclonal antibodies to leucocyte function associated molecule-1 and intracellular adhesion molecule-1. Journal of Immunology 150: 655-663.
  124. Teunissen, M.B.M., Rongen, H.A.H. & Bos, J.D. (1994). Function of adhesion molecules LFA-3 and ICAM-1 on human epidermal Langerhans cells in antigen specific T-cell activation. Journal of Immunology 152: 3400-3409.
  125. Aiba, S., Nakagawa, S., Ozawa, H., Miyaka, K., Ygita, H. & Tagami, H. (1993). Upregulation of a4-integrin on activated Langerhans cells: analysis of adhesion molecules on Langerhans cells relating to their migration from skin to draining lymph nodes. Journal of Investigative Dermatology 100: 143-147.
  126. Koszik, F., Strunk, D., Simonitch, I., Picker, L.J., Stingl, G. & Payer, E. (1994). Expression of monoclonal antibody HECA-452 defined E-selectin ligands on Langerhans cells in normal and diseased skin. Journal of Investigative Dermatology 102: 773-780.
  127. Kobayashi, Y., Staquet, M.J., Dezutter-Dambuyant, C. & Schmitt, D. (1994). Development of motility of Langerhans cells through extracellular matrix by in vitro hapten contact. European Journal of Immunology 24: 2254-2257.
  128. Tammi, R. & Maibach, H.I. (1987). Skin organ culture: why? International Journal of Dermatology 26: 150-160.
  129. Messadi, D.V., Pober, J.S. & Murphy, G.F. (1988). Effects of recombinant y-interferon on HLA-DR and DQ in short term skin organ culture. Laboratory Investigations 58: 61-67.
  130. Le Poole, I.C., Das, P.K., Krieg, S.R., Mekkes, J.R. & Westerhof, W. (1991). Organotypic culture of human skin for studying wound healing. Wounds 3: 102-110.
  131. Rambukkana, A., Bos, J.D., Irik, D., Menko, W., Kapsenberg, M.L. & Das, P.K. (1994). In situ behaviour of human Langerhans cells in skin organ culture. Laboratory Investigations 4: 521-531.
  132. Rambukkana, A., Pistoor, F.H.M., Irik, D., Menko, W., Bos, J.D., Kapsenberg, M.L. & Das, P.K. (1995). Human skin organ culture as an alternative to animal model to study the allergenicity of low molecular weight chemical compounds/haptens. In The World Congress on Alternatives and Animal Use in the Life Sciences: Education, Research, Testing. Alternative Methods in Toxicology and the Life Sciences, Volume 11 (ed. A. Goldberg & L.F.M. van Zutphen), pp. 311-316. New York: Mary Ann Liebert.
  133. Rambukkana, A., Pistoor, F.H.M., Bos, J.D., Kapsenberg, M.L. & Das, P.K. (1996). Effects of contact allergens on human Langerhans cells in skin organ culture: migration, modulation of cell surface molecules and early expression of IL-1β protein. Laboratory Investigations 79: 422-436.
  134. Pistoor, F.H.M., Rambukkana, A., Kroezen, M., Lepoittevin, J.P., Bos, J.D., Kapsenberg, M.L. & Das, P.K. A novel predictive assay for contact allergens using skin explant cultures. American Journal of Pathology, in press.
  135. Pineau, N., Dossou, K.G. & de Silva, O. (1994). Antigen presentation by murine epidermal cells in vitro. In In Vitro Skin Toxicology. Irritation, Phototoxicity, Sensitisation (ed. A. Rougier, A.M. Goldberg & H.I. Maibach), pp. 325-331. New York: Mary Ann Liebert.
  136. Krasteva, M., Péguet-Navarro, J., Moulon, C., Courtellemont, P., Redziniak, G. & Schmitt, D. (1996). In vitro primary sensitisation of hapten-specific T cells by cultured human epidermal Langerhans cells - a screening predictive assay for contact sensitizers. Clinical and Experimental Allergy 26: 563-570.
  137. McKenzie, R.C. & Sauder, D.N. (1990). The role of keratinocyte cytokines in inflammation and immunity. Journal of Investigative Dermatology 95: 105S-107S.
  138. Kupper, T.S. (1990). Immune and inflammatory processes in cutaneous tissues. Mechanisms and speculations. Journal of Clinical Investigations 86: 1783-1789.
  139. Barker, J.N.W.N., Mitra, R.S., Griffiths, C.E.M., Dixit, V.M., & Nickoloff, B.J. (1991). Keratinocytes as initiators of inflammation. Lancet 337: 211-214.
  140. Matsue, H., Cruz, P., Bergstresser, P.R. & Takashima, A. (1992). Cytokine expression by epidermal cell subpopulations. Journal of Investigative Dermatology 99: 42S-54S.
  141. Kimber, I. (1994). Cytokines and regulation of allergic sensitization to chemicals. Toxicology 93: 1-11.
  142. Aragane, Y., Riemann, H., Bhardwaj, R.S., Schwarz, A., Sawada, Y., Yamasa, H., Luger, T.A., Kubinm, M., Trincheri, G. & Schwarz, T. (1994). IL-12 is expressed and released by human keratinocytes and epidermoid carcinoma cell lines. Journal of Immunology 153: 5366-5372.
  143. Howie, S.E.M., Aldridge, R.D., McVittie, E., Forsey, R.J., Sands, C. & Hunter, J.A.A. (1995). Epidermal keratinocyte production of interferon-y immunoreactive protein and mRNA is an early event in allergic contact dermatitis. British Journal of Dermatology 132: 657.
  144. Ponec, M. (1992). In vitro cultured human skin cells as alternatives to animals for skin irritancy screening. International Journal of Cosmetology 14: 245-264.
  145. von den Driesch, P., Fartasch, M., Hunter, A. & Ponec, M. (1995). Expression of integrin receptors and ICAM-1 on keratinocytes in vivo and in an in vitro reconstructed epidermis: effect of sodium dodecyl sulfate. Archives of Dermatological Research 287: 249-253.
  146. Herzinger, T., Korting, H.C. & Maibach, H.I. (1995). Assessment of cutaneous and ocular irritancy: a decade of research on alternatives to animal experimentation. Fundamental and Applied Toxicology 24: 29-41.
  147. Ponec, M. (1995). Reconstructed epidermis in vitro: an alternative to animal testing. ATLA 23: 97-110.
  148. Ponec, M. (1996). In vitro models. In Irritant Contact Dermatitis Syndrome (ed. P. van der Valk & H.I. Maibach), pp. 335-342. Boca Raton, Florida: CRC Press.
  149. Ponec, M. & Kempenaar, J. (1995). Use of human skin recombinants as an in vitro model for testing the irritation potential of cutaneous irritants. Skin Pharmacology 8: 49-59.
  150. Basketter, D.A., Scholes, E.W., Chamberlain, M. & Barratt, M.D. (1995). An alternative strategy to the use of guinea pigs for the identification of skin sensitisation hazard. Food and Chemical Toxicology 33: 1051-1056.
  151. Basketter, D.A. & Scholes, E.W. (1992). Comparison of the local lymph node assay with the guinea pig maximization test for the detection of a range of contact allergens. Food and Chemical Toxicology 60: 65-69.
  152. Kimber, I. & Dearman, R.J. (1993). Approaches to the identification and classification of chemical allergens in mice. Journal of Pharmacological and Toxicological Methods 29: 11-16.
  153. Scholes, E.W., Basketter, D.A., Sarll, A.E., Kimber, I., Evans, C.D., Miller, K., Robbins, M.C., Harrison, P.T.C. & Waite, S.J. (1992). The local lymph node assay: results of a final interlaboratory validation under field conditions. Journal of Applied Toxicology 12: 217-222.
  154. Kligman, A.M. (1966). The identification of contact allergens by human assay. Journal of Investigative Dermatology 47: 393-409.
  155. Wahlberg, J.E. & Boman, A. (1986). Contact sensitivity to arsenical compounds. Clinical and experimental studies. Dermatosen Beruf Umwelf 34: 10-12.