The trilogy of human musk receptors: linking receptor activation, genotype, and sensory perception.

The scent of musk plays a unique role in the history of perfumery. Musk odorants comprise 6 diverse chemical classes and perception differences in strength and quality among human panelists have long puzzled the field of olfaction research. Three odorant receptors (OR) had recently been described for musk odorants: OR5AN1, OR1N2, and OR5A2. High functional expression of the difficult-to-express human OR5A2 was achieved by a modification of the C-terminal domain and the link between sensory perception and receptor activation for the trilogy of these receptors and their key genetic variants was investigated: All 3 receptors detect only musky smelling compounds among 440 commercial fragrance compounds. OR5A2 is the key receptor for the classes of polycyclic and linear musks and for most macrocylic lactones. A single P172L substitution reduces the sensitivity of OR5A2 by around 50-fold. In parallel, human panelists homozygous for this mutation have around 40-60-fold higher sensory detection threshold for selective OR5A2 ligands. For macrocyclic lactones, OR5A2 could further be proven as the key OR by a strong correlation between in vitro activation and the sensory detection threshold in vivo. OR5AN1 is the dominant receptor for the perception of macrocyclic ketones such as muscone and some nitromusks, as panelists with a mutant OR5A2 are still equally sensitive to these ligands. Finally, OR1N2 appears to be an additional receptor involved in the perception of the natural (E)-ambrettolide. This study for the first time links OR activation to sensory perception and genetic polymorphisms for this unique class of odorants.

Avoiding a reproducibility crisis in regulatory toxicology-on the fundamental role of ring trials.

The ongoing transition from chemical hazard and risk assessment based on animal studies to assessment relying mostly on non-animal data, requires a multitude of novel experimental methods, and this means that guidance on the validation and standardisation of test methods intended for international applicability and acceptance, needs to be updated. These so-called new approach methodologies (NAMs) must be applicable to the chemical regulatory domain and provide reliable data which are relevant to hazard and risk assessment. Confidence in and use of NAMs will depend on their reliability and relevance, and both are thoroughly assessed by validation. Validation is, however, a time- and resource-demanding process. As updates on validation guidance are conducted, the valuable components must be kept: Reliable data are and will remain fundamental. In 2016, the scientific community was made aware of the general crisis in scientific reproducibility-validated methods must not fall into this. In this commentary, we emphasize the central importance of ring trials in the validation of experimental methods. Ring trials are sometimes considered to be a major hold-up with little value added to the validation. Here, we clarify that ring trials are indispensable to demonstrate the robustness and reproducibility of a new method. Further, that methods do fail in method transfer and ring trials due to different stumbling blocks, but these provide learnings to ensure the robustness of new methods. At the same time, we identify what it would take to perform ring trials more efficiently, and how ring trials fit into the much-needed update to the guidance on the validation of NAMs.

In vitro to in vivo extrapolation to derive a metabolism factor for estimating the aggregate exposure to salicylic acid after dermal exposure of its esters.

As part of the safety assessment of salicylate esters in cosmetics, we developed a metabolism factor based on in vitro to in vivo extrapolation (IVIVE) to provide a better estimation of the aggregate internal exposure to the common metabolite, salicylic acid. Optimal incubation conditions using human liver S9 were identified before measuring salicylic acid formation from 31 substances. Four control substances, not defined as salicylic esters but which could be mistaken as such due to their nomenclature, did not form salicylic acid. For the remaining substances, higher in vitro intrinsic clearance (CLint, in vitro) values generally correlated with lower LogP values. A "High-Throughput Pharmacokinetic" (HTPK) model was used to extrapolate CLint, in vitro values to human in vivo clearance and half-lives. The latter were used to calculate the percentage of substance metabolised to salicylic acid in 24 h in vivo following human exposure to the ester, i.e. the "metabolism factor". The IVIVE model correctly reproduced the observed elimination rate of 3 substances using in silico or in vitro input parameters. For other substances, in silico only-based predictions generally resulted in lower metabolism factors than when in vitro values for plasma binding and liver S9 CLint, in vitro were used. Therefore, in vitro data input provides the more conservative metabolism factors compared to those derived using on in silico input. In conclusion, these results indicate that not all substances contribute equally (or at all) to the systemic exposure to salicylic acid. Therefore, we propose a realistic metabolism correction factor by which the potential contribution of salicylate esters to the aggregate consumer exposure to salicylic acid from cosmetic use can be estimated.

Bioconcentration Assessment in Fish Based on In Vitro Intrinsic Clearance: Predictivity of an Empirical Model Compared to In Vitro-In Vivo Extrapolation Models.

To estimate the bioconcentration factor (BCF), the in vitro intrinsic clearance (CLIN VITRO,INT) from rainbow trout liver S9 fractions (RT-S9) can be applied to in vitro-in vivo extrapolation (IVIVE) models, yet uncertainties remain in model parameterization. An alternative model approach is evaluated: a regression model was built in the form log BCF = a × log Kow + b × log CLIN VITRO,INT. The coefficients a and b were fitted based on a training set of 40 chemicals. A high robustness of the coefficients and good accuracy of BCF prediction were found on independent datasets of neutral organic chemicals (measured log Kow 3.3-6.2). BCF predictions were similar to or in better agreement with in vivo BCFs compared to IVIVE models (2.4- to 2.9- vs 2.8- to 3.6-fold misprediction) for training and test sets. Species-matched models (trout, carp) did not result in improvements. This study presents the largest dataset on CLIN VITRO,INT and BCFs to assess predictivity of the RT-S9 assay. The robustness of the regression statistics on different datasets and the high statistical weight of the CLIN VITRO,INT term illustrate the predictive power of the RT-S9 assay as an important step toward regulatory acceptance to replace animal experiments.

ECHA ARN documents: chemical grouping without a toxicological rationale.

The EU chemical strategy for sustainability (CSS) plans to use chemical grouping to "prioritise (…) substances for restrictions for all uses through grouping, instead of regulating them one by one". Thus, toxicological grouping will become a key tool used by regulatory authorities in Europe. Over the last 2 years, ECHA has published a high number of documents labelled "Assessment of Regulatory Needs (ARN)" which are based on groups of chemicals based on structural considerations. The ARN documents are legally non-binding, yet they present the public impression of a conclusion about restrictions for groups or sub-groups of chemicals and hence may set a precedent for further binding actions. ECHA has set out definitions on what is considered a group in REACH Annex XI. However, as shown in this commentary based on five examples, the ARN do not follow these principles and propose toxicological groupings without taking into consideration mode of action and the toxicological information on the chemicals. Given the emphasis on grouping projected by the CSS, the groupings in the ARN set an unfortunate precedent on what a toxicological group means and they do not follow clear scientific standards or established toxicological principles. They also lead to a public image of guilt by association for chemicals, without any recourse for registrants to establish the scientific basis for their safe use, as presented within REACH registrations.

Towards the mechanism of spermatotoxicity of p-tert-butyl-alpha-methylhydrocinnamic aldehyde: inhibition of late stage ex-vivo spermatogenesis in rat seminiferous tubule cultures by para-tert-butyl- benzoic acid.

Molecules metabolized to para-tert-butyl-benzoic acid (p-TBBA) affect male reproduction in rats through effects on spermatogenesis. This toxicity is specific to p-TBBA and not observed in meta-substituted analogues. The underlying mode of action was evaluated by comparing effects of p-TBBA and the position isomer m-TBBA (2-50 µM) in an ex vivo 3D primary seminiferous tubule cell culture system from juvenile Sprague Dawley rats (Bio-AlteR®). Treated cultures were evaluated for CoA-conjugate formation, cytotoxicity, blood-testis barrier functionality and different germ cell populations to assess effects on spermatogenesis. In addition, an evaluation of the metabolome of treated cultures was performed by using MxP® Broad Profiling via a LC-MS/MS and GC-MS platform. Para-TBBA decreased germ cell populations of late stages of spermatogenesis and led to the formation of CoA-conjugates in the ex vivo tissue. In addition, p-TBBA had a pronounced effect on the metabolome by affecting lipid balance and other CoA-dependent pathways contributing to energy production and the redox system. Meta-TBBA did not affect germ cell populations and no m-TBBA related CoA-conjugates were detectable. The metabolic profile of m-TBBA treated cells was comparable to vehicle control treated cultures, indicating that formation of CoA-conjugates, inhibition of spermatogenesis, and effects on the metabolome are mechanistically linked events. Thus, for this specific chemical group an adverse outcome pathway can be postulated, including the formation of benzoic acid metabolites, accumulation of CoA-conjugates to a certain threshold and CoA depletion, which affects the metabolic and lipid profile and leads to tissue specific effects with impaired functionalities such as spermatogenesis.

Reduced specificity for the local lymph node assay for lipophilic chemicals: Implications for the validation of new approach methods for skin sensitization.

Meaningful and accurate reference data are crucial for the validation of New Approach Methodologies (NAMs) in toxicology. For skin sensitization, multiple reference datasets are available including human patch test data, guinea pig data and data from the mouse local lymph node assay (LLNA). When assessed against the LLNA, a reduced sensitivity has been reported for in vitro and in chemico assays for lipophilic chemicals with a LogP ≥3.5, resulting in reliability restrictions within the h-CLAT OECD test guideline. Here we address the question of whether LLNA data are an appropriate reference for chemicals in this physicochemical range. Analysis of LLNA vs human reference data indicates that the false-discovery rate of the LLNA is significantly higher for chemicals with LogP ≥3.5. We present a mechanistic hypothesis whereby irritation caused by testing lipophilic chemicals at high test doses leads to unspecific cell proliferation. The accompanying analysis indicates that for lipophilic chemicals with negative calls in in vitro and in chemico assays, resorting to the LLNA is not necessarily a better option. These results indicate that the validation of NAMs in this particular LogP range should be based on a more holistic evaluation of the reference data and not solely upon LLNA data.

Reference Chemical Potency List (RCPL): A new tool for evaluating the accuracy of skin sensitisation potency measurements by New Approach Methodologies (NAMs).

Considerable progress has been made in the design of New Approach Methodologies (NAMs) for the hazard identification of skin sensitising chemicals. However, effective risk assessment requires accurate measurement of sensitising potency, and this has proven more difficult to achieve without recourse to animal tests. One important requirement for the development and adoption of novel approaches for this purpose is the availability of reliable databases for determining the accuracy with which sensitising potency can be predicted. Some previous approaches have relied on comparisons with potency estimates based on either human or animal (local lymph node assay) data. In contrast, we here describe the development of a carefully curated Reference Chemical Potency List (RCPL) which is based on consideration of the best available human and animal data. The RCPL is comprised of 33 readily available chemicals that span a wide range of chemistry and sensitising potency, and contain examples of both direct and indirect (pre- and pro-) haptens. For each chemical a potency value (PV) was derived, and chemicals ranked according to PV without the use of potency categories. It is proposed that the RCPL provides an effective resource for assessment of the accuracy with which NAMs can measure skin sensitising potency.

Nature-derived epoxy resins: Synthesis, allergenicity, and thermosetting properties of pinoresinol diglycidyl ether.

We describe a novel nature-derived epoxy resin monomer (ERM) derived from the plant lignan pinoresinol. Epoxy resins are thermosetting materials in global usage owing to their excellent technical properties such as flexibility and durability. However, their adverse health effects are often not considered and affect users of epoxy resins worldwide. Components of epoxy resin systems are strong skin sensitizers and cause allergic contact dermatitis. The reported prevalence attributable to epoxy chemicals is between 11.7 and 12.5% of all cases of occupational allergic contact dermatitis. We are committed to developing epoxy resins with reduced allergenic effect, while maintaining their excellent properties. The novel ERM, pinoresinol diglycidyl ether (PinoDGE), was synthesized in one step from pinoresinol and epichlorohydrin in 88% yield. It was not classified as a skin sensitizer in the in vivo local lymph node assay, at concentrations up to 0.17 m, as it did not cause a stimulation index >3 compared to control. Pinoresinol diglycidyl ether reacted with the model peptide AcPHCKRM in a reactivity assay and was predicted to be a skin sensitizer in the KeratinoSens assay. Preliminary cross-linking studies indicate that it has promising properties compared to commercially used ERMs. Pinoresinol diglycidyl ether could be seen as a lead compound for further development of alternative ERMs with a better safety profile based on natural and renewable sources for construction of epoxy resin polymers.

Scientific discrepancies in European regulatory proposals on endocrine disruptors-REACH regulation quo vadis?

The EU chemical strategy for sustainability places a high focus on endocrine-disrupting chemicals (ED), the importance of their identification with increased testing and a ban in consumer products by a generic approach. It is assumed that for ED no threshold and hence no safe dose exists, leading to this generic approach. This view appears to be linked to the claim that for ED 'low-dose non-monotonic dose response' (low-dose NMDR) effects are observed. Without this hypothesis, there are no scientific reasons why classical risk assessment cannot be applied to the ED mode-of-action. Thus, whether for ED low-dose NMDR effects are considered a reproducible scientific fact by European authorities is Gretchen's question in this politicized field. Recent documents by the SCCS, EFSA and ECHA reviewed herein illustrate the diverging views within European scientific bodies on this issue. Furthermore, ED researchers never replicated findings on low-dose NMDR in blinded inter-laboratory experiments and the CLARITY-BPA core studies could not find evidence for reproducible NMDR for BPA. ECHA proposes a battery of in vitro tests to test all chemicals for ED properties. However, these tests were never validated for relevance and their high positivity rate could lead to increased follow-up animal testing. Based on (i) lack of reproducibility data for low-dose NMDR, (ii) diverging views within European authorities on NMDR and (iii) lack of fully validated in vitro test methods it might be premature to fast-track the wide-ranging changes in the regulatory landscape proposed by the authorities ultimately leading to drastically increased animal testing.

Examining Uncertainty in In Vitro-In Vivo Extrapolation Applied in Fish Bioconcentration Models.

In vitro biotransformation rates were determined for 30 chemicals, mostly fragrance ingredients, using trout liver S9 fractions (RT-S9) and incorporated into in vitro-in vivo extrapolation (IVIVE) models to predict bioconcentration factors (BCFs). Predicted BCFs were compared against empirical BCFs to explore potential major uncertainties involved in the in vitro methods and IVIVE models: (i) in vitro chemical test concentrations; (ii) different gill uptake rate constant calculations (k1); (iii) protein binding (different calculations and measurement of the fraction of unbound chemical, fU); (iv) species differences; and (v) extrahepatic biotransformation. Predicted BCFs were within 0.5 log units for 44% of the chemicals compared to empirical BCFs, whereas 56% were overpredicted by >0.5 log units. This trend of overprediction was reduced by alternative k1 calculations to 32% of chemicals being overpredicted. Moreover, hepatic in vitro rates scaled to whole body biotransformation rates (kB) were compared against in vivo kB estimates. In vivo kB was underestimated for 79% of the chemicals. Neither lowering the test concentration, nor incorporation of new measured fU values, nor species matching avoided the tendency to overpredict BCFs indicating that further improvements to the IVIVE models are needed or extrahepatic biotransformation plays an underestimated role.

Benzoyl-CoA conjugate accumulation as an initiating event for male reprotoxic effects in the rat? Structure-activity analysis, species specificity, and in vivo relevance.

A number of para-substituted benzoic acids (p-BA) and chemicals metabolized to p-BA have been found to confer adverse effects in male rats on sperm viability, motility, and morphology. These effects are putatively associated with the metabolism of p-BA to toxic intermediates. We had shown that p-BA lead to accumulation of high levels of p-alkyl-benzoyl-CoA conjugates in plated primary rat hepatocytes. Here we further investigated the relevance of this metabolic pathway for the reprotoxic effects in rats and rabbits. We extended the structure-activity relationship to a set of 19 chemicals (nine reprotoxic and ten non-reprotoxic) and confirmed a very strong correlation between p-alkyl-benzoyl-CoA accumulation in rat hepatocytes and the toxic outcome. Species specificity was probed by comparing rat, rabbit and human hepatocytes, and p-benzoyl-CoA accumulation was found to be specific to the rat hepatocytes, not occurring in human hepatocytes. There was also very limited accumulation in hepatocytes from rabbits that are a non-responder species in in vivo studies. Tissues of rats treated with 3-(4-isopropylphenyl)-2-methylpropanal were analysed and p-isopropyl-benzoyl-CoA conjugates were detected in the liver and in the testes in animals at toxic doses indicating that the metabolism observed in vitro is relevant to the in vivo situation and the critical metabolite does also occur in the reproductive tissue. These multiple lines of evidence further support benzoyl-CoA accumulation as a key initiating event for a specific group of male reproductive toxicants, and indicate a species-specific effect in the rat.

Replacing the refinement for skin sensitization testing: Considerations to the implementation of adverse outcome pathway (AOP)-based defined approaches (DA) in OECD guidelines.

While single non-animal methods have been adopted in OECD test guidelines, combinations of methods (so called defined approaches, DA) are not. Hardly any animal study can be replaced by a single non-animal method, rather DA are needed. The OECD published the Adverse Outcome Pathway (AOP) on skin sensitization in 2012 and is currently discussing the implementation of DA into a guideline. Obviously, it takes thorough considerations and evaluations to validate such DA. Currently we see four preconditions for a proper and expedient implementation of DA in a guideline: (i) The reference data should be selected to allow meaningful evaluations and must not replicate the limitations of the murine local lymph node assay (LLNA) (ii) Methods and prediction models should be validated before they are used in an OECD-adopted DA, (iii) An OECD-adopted DA should follow the respective AOP and (iv) acknowledge regulatory needs and successful toxicological practice. These points still need to be considered in the current discussion at the OECD. A guideline for skin sensitization DA is setting the scene for regulatory acceptance of all new approaches (for any toxicological endpoint) in the future. In this commentary, we are expounding these preconditions to allow a scientifically valid and sustainable application of modern (no-animal) approaches in regulatory toxicology.

PeBiToSens™: A Platform for PBT Screening of Fragrance Ingredients Without Animal Testing.

The determination of persistence (P), bioaccumulation (B) and toxicity (T) plays a central role in the environmental assessment of chemicals. Persistence is typically evaluated via standard microbial biodegradation tests. Bioaccumulation refers to the accumulation of chemicals in organisms and is usually assessed in fish exposed to the test chemical. Toxicity is determined at three trophic levels, with fish toxicity as the highest trophic level assessed. Thus, animal tests are classically needed for both B and T assessment. In vitro systems based on fish liver cells or liver S9 fractions ('RT-S9 assay') have been recently adopted by OECD to measure the biotransformation rates for the chemicals for B assessment. Biotransformation drives clearance from the body and reduces bioaccumulation. For T assessment, an assay based on in vitro toxicity on fish gill cells has been established ('RTgill-W1 assay'). Here we summarize our findings indicating that these tests are highly predictive for fragrance ingredients, and show with two case studies of our latest new registered substances how we apply these tests in particular during development and also for chemical registration. This platform of tests (PeBiToSens™) could fully replace animal tests in ecotoxicological assessment and is key in the Givaudan Safe by Design™ approach to develop safer and environmentally compatible novel fragrance ingredients.

A review of substances found positive in 1 of 3 in vitro tests for skin sensitization.

There has been significant progress in recent years in the development and application of alternative methods for assessing the skin sensitization potential of chemicals. The pathways involved in skin sensitization have been described in an OECD adverse outcome pathway (AOP). To date, a single non-animal test method is not sufficient to address this AOP so numerous approaches involving the use of 2 or more assays are being evaluated for their performance. The 2 out of 3 approach is a simple approach that has demonstrated very good sensitivity, specificity and overall accuracy numbers for predicting the skin sensitization potential of chemicals. Chemicals with at least two positive results in tests addressing Key events 1-3 are predicted sensitizers, while chemicals with none or only one positive outcome are predicted non-sensitizers. In this report we have thoroughly reviewed the discordant results of 29 chemicals with 1 out of 3 positive results to understand better what led to the results observed and how this information might impact our hazard assessments of these chemicals. We initially categorized each chemical using a weight of evidence approach as positive, negative or indeterminate based on review of available human and animal data as well as what skin sensitization alerts were triggered using two versions of OECD Toolbox and DEREK Nexus. We determined that 4 of the 29 chemicals should be classified as indeterminate and not included in analysis of method performance based on insufficient, borderline and/or conflicting data to confidently categorized the chemicals as allergens or non-allergens. Of the 29 chemicals included in this analysis, 17 were classified as negative and would be correctly identified using a 2 out of 3 approach while 8 chemicals were classified as positive in vivo and would be false-negative with this approach. For some of these chemicals, the outcomes observed can be explained by in vitro borderline results (13 chemicals) or in some instances there is mechanistic understanding of why a chemical is positive or negative in a particular assay (9 chemicals). Thus, when comparing the performance of different defined approaches, one should attempt to only include chemicals which demonstrate clear evidence to be categorize as allergens or non-allergens. Finally, when interpreting the results obtained for an individual unknown chemical it is critical that the in vitro skin sensitization data is reviewed critically and there is a good understanding of the variance and applicability domain limitations for each assay being used.

Effect of Fluorination on Skin Sensitization Potential and Fragrant Properties of Cinnamyl Compounds.

A series of three α- and three ß-fluorinated representatives of the family of cinnamate-derived odorants (cinnamaldehyde (1), cinnamyl alcohol (2), and ethyl cinnamate (3)) as used as fragrance ingredients is described. Olfactive evaluation shows that the fluorinated compounds exhibit a similar odor profile to their parent compounds, but the olfactive detection thresholds are clearly higher. In vitro evaluation of the skin sensitizing properties with three different assays indicates that α-fluorination of Michael acceptor systems 1 and 3 slightly improves the skin sensitization profile. α-Fluorocinnamyl alcohol 2b is a weaker skin sensitizer than cinnamyl alcohol 2a by in vitro tests and the fluorinated product drops below the sensitization threshold of the KeratinoSens® assay. On the other hand, ß-fluorination of compounds 1 - 3 results in highly reactive products which display a worsened in vitro skin sensitization profile.

Reaction Chemistry to Characterize the Molecular Initiating Event in Skin Sensitization: A Journey to Be Continued.

The publication by the OECD of the adverse outcome pathway (AOP) for skin sensitization has accelerated the development and validation of mechanistic tests and testing strategies to assess the potential of new molecules to trigger skin allergies. The molecular initiating event (MIE) in the AOP is reaction with skin peptides/proteins. It is followed by a number of cellular events. Currently, only one in chemico test to characterize the MIE was proposed to and underwent adoption by the OECD, while two cell-based assays have completed the process. A multitude of further cellular assays is currently in the validation pipeline, but no further reactivity test has gone through full standardization. Here, we review data on in chemico methods, identify gaps, and discuss how these methods can be improved to better characterize the MIE and to become even more informative. We focus on the importance of kinetics, the information gained from studying adduct formation, and the difficulties posed by side reactions such as peptide oxidation. We then highlight mechanistic insights from reaction chemistry: the relative contribution of different target nucleophiles, the migration of amino acid modifications, and the potential of peptide-cross-linking. We illustrate in a case study how kinetic in chemico methods might have been used to better predict the risk of three preservatives, which have led to serious epidemics of contact dermatitis. In a case study on Michael acceptors, we show the impact of additional substituents around the electrophilic olefin moiety on reactivity, and we highlight the shortcomings which current in silico methods to predict reaction chemistry still have, illustrating the need for experimental in chemico data to improve such models. Finally, based on the information reviewed and the presented case studies, a strong argument is made to continue the journey of developing nonredundant, informative in chemico methods, and not to solely focus on new cell-based methods to further populate the AOP for skin sensitization.

Nrf2 activation as a key event triggered by skin sensitisers: The development of the stable KeratinoSens reporter gene assay.

The 21st century paradigm for toxicology and the adverse outcome pathway concept envisage a future toxicology largely based on mechanistic in vitro assays and relying mainly on cellular models. In the skin sensitisation field, this concept was not intuitive at the beginning. Given the high structural diversity of skin sensitising molecules, classical receptor binding as the molecular initiating event in a cell-based assay could be excluded from the start, leaving the question of how cells could sense potential skin sensitising chemicals and be able to differentiate them from non-sensitisers. When we entered this field in 2006, we realised that, in another emerging field of toxicology, detailed work on the antioxidant/electrophile sensing pathway Keap1/Nrf2/ARE was being performed. We postulated that, based on their intrinsic electrophilicity, a large structural variety of skin sensitisers would activate this pathway. This was demonstrated in a preliminary pilot study with an existing, breast cancer-derived reporter cell line. Broader confirmation of this initial hypothesis then came from a multitude of genome-wide studies, in which sensitiser-induced changes to the transcriptome were investigated. The results showed that this regulatory pathway is indeed the most common regulatory pathway activated by sensitisers at the gene expression level, and the underlying event in keratinocytes has become formalised as a Key Event in the Organisation for Economic Co-operation and Development (OECD) Adverse Outcome Pathway for sensitisation. These studies led to the development of the KeratinoSens® assay, which became the first cell-based in vitro test for skin sensitisation to be endorsed by a European Union Reference Laboratory for Alternatives to Animal Testing (EURL ECVAM) statement and an OECD Test Guideline. More recently, a number of studies have further developed this approach into 3-D skin models. Here, we review the underlying mechanism and the development of the KeratinoSens assay. We also present data on the stability of the assay over time, which is a key requirement for a cell-based biological assay to be endorsed in a regulatory context.

Dual regulation of skin sensitizer-induced HMOX1 expression by Bach1 and Nrf2: Comparison to regulation of the AKR1C2-ARE element in the KeratinoSens cell line.

Heme oxygenase (decycling) 1 (HMOX1) is the most consistently found genetic marker induced by skin sensitizers. HMOX1 is often referred to as typical gene regulated by nuclear factor erythroid 2-related factor 2 (Nrf2), however, it is also regulated by other DNA-binding factors, including BTB and CNC homolog 1 (Bach1). The KeratinoSens™ assay is the first validated in vitro assay for sensitizers that measures gene induction. It is based on luciferase expression regulated by the antioxidant response element (ARE) of the aldoketoreductase 1C2 (AKR1C2) gene. Luciferase upregulation is dependent on Nrf2, while HMOX1 upregulation is only partially Nrf2-dependent. Thus, sensitizer-dependent activation of HMOX1 may integrate multiple signals thereby providing additional information. We constructed reporter cell lines containing the full HMOX1 regulatory region or the HMOX1-ARE sequence and compared them with the construct containing the AKR1C2-ARE sequence. Induction of the AKR1C2-ARE depends on Nrf2, but not on the repressor Bach1. Results obtained with HMOX1-ARE and the full HMOX1 promoter indicate that, within the HMOX1 promoter, the HMOX1-ARE is sufficient to explain the induction by sensitizers and that (i) inhibiting Bach1 leads to strong basal expression, (ii) fold-induction by sensitizers above this level is reduced in the absence of Bach1 and (iii) these constructs are less dependent on Nrf2 as compared to the AKR1C2-ARE. Nevertheless, congruent dose response curves for luciferase activity were obtained with all constructs. Thus, while sensitizer-induced HMOX1 activation is dependent on Nrf2 and Bach1, all constructs give identical information for the in vitro prediction of the sensitization potential.