Quantitative assessment of sensitizing potency using a dose-response adaptation of GARDskin

Nature Scientific Reports 11, 18904 (2021), https://doi.org/10.1038/s41598-021-98247-7

Robin Gradin, Andy Forreryd, Ulrika Mattson, Anders Jerre, Henrik Johansson


Hundreds of chemicals have been identified as skin sensitizers. These are chemicals that possess the ability to induce hypersensitivity reactions in humans, giving rise to a condition termed allergic contact dermatitis. The capacity to limit hazardous exposure to such chemicals depends upon the ability to accurately identify and characterize their skin sensitizing potency. This has traditionally been accomplished using animal models, but their widespread use offers challenges from both an ethical and a scientific perspective. Comprehensive efforts have been made by the scientific community to develop new approach methodologies (NAMs) capable of replacing in vivo assays, which have successfully yielded several methods that can identify skin sensitizers. However, there is still a lack of new approaches that can effectively measure skin sensitizing potency. We present a novel methodology for quantitative assessment of skin sensitizing potency, which is founded on the already established protocols of the GARDskin assay. This approach analyses dose-response relationships in the GARDskin assay to identify chemical-specific concentrations that are sufficient to induce a positive response in the assay. We here compare results for 22 skin sensitizers analyzed using this method with both human and LLNA potency reference data and show that the results correlate strongly and significantly with both metrics (rLLNA = 0.81, p = 9.1 × 10–5; rHuman = 0.74, p = 1.5 × 10–3).

In conclusion, the results suggest that the proposed GARDskin dose-response methodology provides a novel non-animal approach for quantitative potency assessment, which could represent an important step towards reducing the need for in vivo experiments.


Key words: GARD, GARDskin, GARDskin Dose-Response, in vitro, sensitization, potency, chemical sensitizers, quantitative risk assessment


Full article
Article on line with open access

The GARDpotency assay for potency-associated subclassification of chemical skin sensitizers – Rationale, method development and ring trial results of predictive performance and reproducibility

Toxicological Sciences, kfaa068, https://doi.org/10.1093/toxsci/kfaa068

Robin Gradin, Angelica Johansson, Andy Forreryd, Emil Aaltonen, Anders Jerre, Olivia Larne, Ulrika Mattson, Henrik Johansson

Proactive identification and characterization of hazards attributable to chemicals are central aspects of risk assessments. Current legislations and trends in predictive toxicology advocate a transition from in vivo methods to non-animal alternatives. For skin sensitization assessment, several OECD validated alternatives exist for hazard identification, but non-animal methods capable of accurately characterizing the risks associated with sensitizing potency are still lacking.

The GARDTM platform utilizes exposure-induced gene expression profiles of a dendritic -like cell line in combination with machine learning to provide hazard classifications for different immunotoxicity endpoints. Recently, a novel genomic biomarker signature displaying promising potency-associated discrimination between weak and strong skin sensitizers was proposed. Here, we present the adaptation of the defined biomarker signature on a gene expression analysis platform suited for routine acquisition, confirm the validity of the proposed biomarkers, and define the GARDTMpotency assay for prediction of skin sensitizer potency. The performance of GARDTMpotency was validated in a blinded ring-trial, in accordance with OECD-guidance documents. The cumulative accuracy was estimated to 88.0% across three laboratories and nine independent experiments. The within-laboratory reproducibility measures ranged between 62.5% and 88.9%, and the between-laboratory reproducibility was estimated to 61.1%. Currently, no direct or systematic cause for the observed inconsistencies between the laboratories have been identified. Further investigations into the sources of introduced variability will potentially allow for increased reproducibility.

In conclusion, the in vitro GARDTMpotency assay constitute a step forward for development of non-animal alternatives for hazard characterization of skin sensitizers.

Key words: GARD, GARDpotency, in vitro, sensitization, potency, chemical sensitizers

Full article
Article on line with open access

An integrated transcriptomic- and proteomic-based approach to evaluate the human skin sensitization potential of glyphosate and its commercial agrochemical formulations

Journal of Proteomics
Available online 30 January 2020, 03647. https://doi.org/10.1016/j.jprot.2020.103647

Tim Lindberg, Renato Ivan de Ávila, Kathrin S. Zeller, Fredrik Levander, Dennis Eriksson, Aakash Chawaded, Malin Lindstedt


  • Pure glyphosate was classified as a non-sensitizer using in vitro assessment.
  • POEA, POEA+glyphosate mixture and formulations were identified as skin sensitizers.
  • MS analysis identified protein groups related to immunologically relevant events.
  • Autophagy may be involved in the agrochemical materials-induced DC responses.

We investigated the skin sensitization hazard of glyphosate, the surfactant polyethylated tallow amine (POEA) and two commercial glyphosate-containing formulations using different omics-technologies based on a human dendritic cell (DC)-like cell line. First, the GARD™skin assay, investigating changes in the expression of 200 transcripts upon cell exposure to xenobiotics, was used for skin sensitization prediction. POEA and the formulations were classified as skin sensitizers while glyphosate alone was classified as a non-sensitizer. Interestingly, the mixture of POEA together with glyphosate displayed a similar sensitizing prediction as POEA alone, indicating that glyphosate likely does not increase the sensitizing capacity when associated with POEA. Moreover, mass spectrometry analysis identified differentially regulated protein groups and predicted molecular pathways based on a proteomic approach in response to cell exposures with glyphosate, POEA and the glyphosate-containing formulations. Based on the protein expression data, predicted pathways were linked to immunologically relevant events and regulated proteins further to cholesterol biosynthesis and homeostasis as well as to autophagy, identifying novel aspects of DC responses after exposure to xenobiotics. In summary, we here present an integrative analysis involving advanced technologies to elucidate the molecular mechanisms behind DC activation in the skin sensitization process triggered by the investigated agrochemical materials.

The use of glyphosate has increased worldwide, and much effort has been made to improve risk assessments and to further elucidate the mechanisms behind any potential human health hazard of this chemical and its agrochemical formulations. In this context, omics-based techniques can provide a multiparametric approach, including several biomarkers, to expand the mechanistic knowledge of xenobiotics-induced toxicity. Based on this, we performed the integration of GARD™skin and proteomic data to elucidate the skin sensitization hazard of POEA, glyphosate and its two commercial mixtures, and to investigate cellular responses more in detail on protein level. The proteomic data indicate the regulation of immune response-related pathways and proteins associated with cholesterol biosynthesis and homeostasis as well as to autophagy, identifying novel aspects of DC responses after exposure to xenobiotics. Therefore, our data show the applicability of a multiparametric integrated approach for the mechanism-based hazard evaluation of xenobiotics, eventually complementing decision making in the holistic risk assessment of chemicals regarding their allergenic potential in humans.

Full article
Article on line ahead of print


Validation of the GARD™skin assay for assessment of chemical skin sensitizers – ring trial results of predictive performance and reproducibility

Toxicological Sciences, kfz108, https://doi.org/10.1093/toxsci/kfz108

Henrik Johansson, Robin Gradin, Angelica Johansson, Els Adriaens, Amber Edwards, Veronika Zuckerstätter, Anders Jerre, Florence Burleson, Helge Gehrke, Erwin Roggen

Proactive identification of chemicals with skin sensitizing properties is a key toxicological endpoint within chemical safety assessment, as required by legislation for registration of chemicals. In order to meet demands of increased animal welfare and facilitate increased testing efficiency also in non-regulatory settings, considerable efforts have been made to develop non-animal approaches to replace current animal testing.

Genomic Allergen Rapid Detection (GARD) is a state-of-the-art technology platform, the most advanced application of which is the assay for assessment of skin sensitizing chemicals, GARDskin. The methodology is based on a dendritic cell (DC)-like cell line, thus mimicking the mechanistic events leading to initiation and modulation of downstream immunological responses. Induced transcriptional changes are measured following exposure to test chemicals, providing a detailed evaluation of cell activation. These changes are associated with the immunological decision-making role of DCs in vivo and include among other phenotypic modifications, up-regulation of co-stimulatory molecules, induction of cellular and oxidative stress pathways and xenobiotic responses and provide a holistic readout of substance-induced DC activation.Here, results from an inter-laboratory ring trial of GARDskin, conducted in compliance with OECD guidance documents and comprising a blinded chemical test set of 28 chemicals, are summarized. The assay was found to be transferable to naïve laboratories, with an inter-laboratory reproducibility of 92.0%. The within-laboratory reproducibility ranged between 82.1-88.9%, while the cumulative predictive accuracy across the three laboratories was 93.8%.It was concluded that GARDskin is a robust and reliable method for the identification of skin sensitizing chemicals and suitable for stand-alone use or as a constituent of integrated testing. These data form the basis for the regulatory validation of GARDskin.
Full article
Article on line with open access 

Evaluation of in vitro testing strategies for hazard assessment of the skin sensitization potential of “real‐life” mixtures: the case of henna‐based hair coloring products containing p‐phenylenediamine

Wiley Online Library, First published: 21 April 2019,  https://doi.org/10.1111/cod.13294.

Renato I. de Ávila Danillo F. M. C. Veloso Gabriel C. Teixeira Thaisângela L. Rodrigues Tim Lindberg Malin Lindstedt Simone G. Fonseca Eliana M. Lima Marize C. Valadares


Allergic contact dermatitis reported to henna‐based hair coloring products (HPs) has been associated with adulteration of henna with p‐phenylenediamine (PPD).

To develop a testing approach based on in vitro techniques that address key events within the skin sensitization adverse outcome pathway to evaluate allergenic potential of HPs.

The following in vitro assays were used to test the sensitizing capacity of hair dye ingredients: micro‐direct peptide reactivity assay (mDPRA); HaCaT keratinocytes‐associated IL‐18 assay; U937 cell line activation test (USENS)/IL‐8 levels; blood monocyte‐derived dendritic cell test; genomic allergen rapid detection (GARD skin). Those techniques with better human concordance were selected to evaluate the allergenic potential of ten HPs.

Contrasting to the label’s information, chromatographic analyses identified PPD in all products. The main henna biomarker, lawsone, was not detected in one of 10 the products. Among the techniques evaluated by testing HDIs, mDPRA, IL‐18 assay, GARD skin and U‐SENS correlated better with human classification (concordances 91.7 to 100%) and were superior to the animal testing (concordance 78.5%). Thus, these assays were used to evaluate HPs, which were classified as skin sensitizers using different two‐out‐of‐three approaches.

Our findings highlight toxicological consequences and risks associated of the undisclosed use of PPD in henna‐based “natural” “real‐life” products.

This article is protected by copyright. All rights reserved.

Skin Sensitization Testing—What’s Next?

Int. J. Mol. Sci. 2019, 20(3), 666; https://doi.org/10.3390/ijms20030666  

Gunilla Grundström and Carl A.K. Borrebaeck


There is an increasing demand for alternative in vitro methods to replace animal testing, and, to succeed, new methods are required to be at least as accurate as existing in vivo tests. However, skin sensitization is a complex process requiring coordinated and tightly regulated interactions between a variety of cells and molecules. Consequently, there is considerable difficulty in reproducing this level of biological complexity in vitro, and as a result the development of non-animal methods has posed a major challenge. However, with the use of a relevant biological system, the high information content of whole genome expression, and comprehensive bioinformatics, assays for most complex biological processes can be achieved. We propose that the Genomic Allergen Rapid Detection (GARD™) assay, developed to create a holistic data-driven in vitro model with high informational content, could be such an example. Based on the genomic expression of a mature human dendritic cell line and state-of-the-art machine learning techniques, GARD™ can today accurately predict skin sensitizers and correctly categorize skin sensitizing potency. Consequently, by utilizing advanced processing tools in combination with high information genomic or proteomic data, we can take the next step toward alternative methods with the same predictive accuracy as today’s in vivo methods—and beyond.


genomics; machine learning; skin sensitization; adverse outcome pathways; next generation in vitro tests

Article online

The use of Genomic Allergen Rapid Detection (GARD) assays to predict the respiratory and skin sensitising potential of e-liquids

Regulatory Toxicology and Pharmacology. Volume 103, April 2019, Pages 158-165 https://doi.org/10.1016/j.yrtph.2019.01.001

Matthew Stevenson, Lukasz Czekala, Liam Simms, Nicole Tschierske, Olivia Larne, Tanvir Walele

Electronic cigarettes (e-cigarettes) are an increasingly popular alternative to combustible tobacco cigarettes among smokers worldwide. A growing body of research indicates that flavours play a critical role in attracting and retaining smokers into the e-cigarette category, directly contributing to declining smoking rates and tobacco harm reduction. The responsible selection and inclusion levels of flavourings in e-liquids must be guided by toxicological principles. Some flavour ingredients, whether natural extracts or synthetic, are known allergens. In this study, we used the Genomic Allergen Rapid Detection (GARD) testing strategy to predict and compare the respiratory and skin sensitising potential of three experimental and two commercial e-liquids. These novel, myeloid cell-based assays use changes in the transcriptional profiles of genomic biomarkers that are collectively relevant for respiratory and skin sensitisation. Our initial results indicate that the GARD assays were able to differentiate and broadly classify e-liquids based on their sensitisation potential, which are defined mixtures. Further studies need to be conducted to assess whether and how these assays could be used for the screening and toxicological assessment of e-liquids to support product development and commercialisation.

Electronic cigarettesSkin sensitisationRespiratory sensitisationIn vitroAlternative methods

Article on line

Is a combination of assays really needed for non-animal prediction of skin sensitization potential? Performance of the GARD™ (Genomic Allergen Rapid Detection) assay in comparison with OECD guideline assays alone and in combination

Regulatory Toxicology and Pharmacology, Volume 98, October 2018, Pages 155-160, https://doi.org/10.1016/j.yrtph.2018.07.014.

David W.Roberts


  • Prediction of skin sensitization potential does not need multiple assays representing Key Events of the AOP.
  • This has been argued on theoretical grounds and is now tested against published data.
  • A single assay, GARD™, can outperform combinations of OECD test guideline assays.

To meet regulatory requirements, and avoid or minimize animal testing, there is a need for non-animal methods to assess the potential of chemicals to cause skin sensitization. It is widely assumed that no one test will be sufficient and that combined data from several assays spanning key events from the adverse outcome pathway will be required. This paper challenges that assumption. The predictive performance of a single assay, the Genomic Allergen Rapid Detection (GARD™) assay, was compared with the performance, singly and in combination, of three formally validated non-animal approaches that appear as OECD test guidelines: the direct peptide reactivity assay (DPRA), the ARE-Nrf2 luciferase test method, and the human cell line activation test (h-CLAT).

It is shown here that GARD™ alone outperforms each of DPRA, ARE-Nrf2 luciferase or h-CLAT, alone or in any combination as a 2 out of 3 strategy, in terms of sensitivity, specificity and accuracy.

Based on the datasets analysed here, the sensitivity and specificity of GARD™ alone are 90–92% and 79–84% (“2 out of 3”, 86% and 76%). Thus, in any situation where the 2 out of 3 strategy is considered adequate, GARD™ alone could be used with equal or better performance.

Skin sensitization potential, Non-animal assays, Integrated testing strategies, GARD™

Full article
Article on line

Predicting skin sensitizers with confidence – Using conformal prediction to determine applicability domain of GARD

Toxicol In Vitro.  Jan 30 2018,  [Epub ahead of print];48:179-187. doi: 10.1016/j.tiv.2018.01.021.

Andy Forreryd, Ulf Norinder, Tim Lindberg, Malin Lindstedt.


GARD – Genomic Allergen Rapid Detection is a cell based alternative to animal testing for identification of skin sensitizers. The assay is based on a biomarker signature comprising 200 genes measured in an in vitro model of dendritic cells following chemical stimulations, and consistently reports predictive performances ~90% for classification of external test sets. Within the field of in vitro skin sensitization testing, definition of applicability domain is often neglected by test developers, and assays are often considered applicable across the entire chemical space. This study complements previous assessments of model performance with an estimate of confidence in individual classifications, as well as a statistically valid determination of the applicability domain for the GARD assay. Conformal prediction was implemented into current GARD protocols, and a large external test dataset (n = 70) was classified at a confidence level of 85%, to generate a valid model with a balanced accuracy of 88%, with none of the tested chemical reactivity domains identified as outside the applicability domain of the assay. In conclusion, results presented in this study complement previously reported predictive performances of GARD with a statistically valid assessment of uncertainty in each individual prediction, thus allowing for classification of skin sensitizers with confidence.

Applicability domain; Conformal prediction; GARD; In vitro assay; Skin sensitization

PMID: 29374571     DOI: 10.1016/j.tiv.2018.01.021

Article on line

An alternative biomarker-based approach for the prediction of proteins known to sensitize the respiratory tract.

Toxicol In Vitro. 2017 Oct 7;46:155-162. doi: 10.1016/j.tiv.2017.09.029.

Zeller KS, Johansson H, Lund TØ, Kristensen NN, Roggen EL, Lindstedt M.



Many natural and industrial proteins are known to have properties that can result in type I hypersensitivity, however, to date, no validated test system exists that can predict the sensitizing potential of these allergens. Thus, the objective of this study was to develop a protocol based on the myeloid cell-based Genomic Allergen Rapid Detection (GARD) assay that can be used to assess and predict the capacity of protein allergens known to induce sensitization in the respiratory tract. Cellular responses induced by eight selected proteins were assessed using transcriptional profiling, flow cytometry and multiplex cytokine analysis. 391 potential biomarkers were identified as a predictive signature and a series of cross-validations supported the validity of the model. These results together with biological pathway analysis of the transcriptomic data indicate that the investigated cell system is able to capture relevant events linked to type I hypersensitization.

Biomarker; Cell-based assay; Dendritic cells; Protein allergens; Respiratory sensitization

Link to article online