Potency classification of skin sensitizers: development and use of in vitro methods Prof. Jean-Pierre Lepoittevin University of Strasbourg, France SCCS WG on methodology Luxemburg, July 1st 2015
Research funding sources Cosmetics Europe Research Institute for Fragrance Material (RIFM) L Oréal Centre National de la Recherche Scientifique (CNRS) Université de Strasbourg
Risk assessment Step 1 Chemical Step 2 Hazard Potency
in vitro models Development of alternative methods to animal testing in silico models Structure activity relationships Hapten Intracellular signaling Epidermis in chemico models hapten-protein Interactions Dendritic cells Keratinocytes Intracellular signaling Transcriptome Maturation Phenotypical modifications Migration T Production of soluble factors TC activation Draining lymphnode in vivo models
Adverse Outcome Pathway (AOP) chemical structures & properties molecular initiating events cellular response Organ response Lymph node Organism response Skin (epidermis) metabolism penetration electrophilic substance covalent interaction with proteins Dendritic cells Keratinocytes MHC presentation Activation of T-cells Proliferation of activated T-cells Inflammation upon challenge Toxicity pathway Mode of action Adverse outcome
Adverse Outcome Pathway (AOP) chemical structures & properties molecular initiating events cellular response Organ response Lymph node Organism response Skin (epidermis) metabolism penetration electrophilic substance covalent interaction with proteins Dendritic cells Keratinocytes MHC presentation Activation of T-cells Proliferation of activated T-cells Inflammation upon challenge In vitro skin absorption (Q)SARs Peptide depletion Adduct formation Activation (Keap-1/Nrf2-ARE Gene expression In vitro T-cell priming / proliferation In silico toxicokinetics Relative reactivity rate Pro-inflammatory mediators Co-stimulatory adhesion molecules
Adverse Outcome Pathway (AOP) chemical structures & properties molecular initiating events KE3 cellular response KE4 Organ response Lymph node Organism response Skin (epidermis) metabolism penetration electrophilic substance KE1 covalent interaction with proteins KE2 Dendritic cells Keratinocytes MHC presentation Activation of T-cells Proliferation of activated T-cells Inflammation upon challenge In vitro skin absorption (Q)SARs Peptide depletion Adduct formation Activation (Keap-1/Nrf2-ARE Gene expression In vitro T-cell priming / proliferation In silico toxicokinetics Relative reactivity rate Pro-inflammatory mediators Co-stimulatory adhesion molecules
in vitro models KE1: hapten-protein interactions in silico models in chemico models Structure activity relationships hapten-protein Interactions Hapten Dendritic cells Epidermis Intracellular signaling Keratinocytes Intracellular signaling Transcriptome Maturation Phenotypical modifications Migration T Production of soluble factors TC activation Draining lymphnode in vivo models
Development of alternative methods to animal testing
KE1: Direct Peptide Reactivity Assay (DPRA) Mechanistic basis: address haptenization of proteins Test system: synthetic heptapeptides containing either Cys or Lys residues End-point: Cys and Lys peptide % depletion Protocol: Test chemical in acetonitrile; Cysteine peptide in phosphate buffer, ph 7.5; Lysine peptide in ammonium acetate, ph 10.2; Test chemical reacted with peptide (10:1 or 50:1) for 24 hours Controls: cinnamic aldehyde (positive)
KE1: Direct Peptide Reactivity Assay (DPRA) 2 synthetic peptides: Pep-Cys Pep-Lys Gerberick F et al. Toxicol Sci, 2004, 81, 332-343
KE1: Direct Peptide Reactivity Assay (DPRA) Test chemical in acetonitrile Cysteine peptide in phosphate buffer, ph 7.5 Lysine peptide in ammonium acetate, ph 10.2 Test chemical reacted with peptide (10:1 or 50:1) for 24 hours Peptide loss monitored by HPLC at 220 nm Un-reacted Peptide Test Chemical Reaction Mixture
KE1:Direct Peptide Reactivity Assay (DPRA) Prediction model: mean percent cysteine and lysine peptide depletion value of 6.38 is used as threshold to discriminate between negative and positive predictions Applicability: not applicable for metals, oxidizers, highly hydrophobic substances, pre- prohaptens
Potential for relation reactivity vs sensitization potential 22.6 % Molecule Pep-Cys > 22.6 % Molecule Molecule 6.4 % > 6.4 % Pep-Lys 42.5 % Pep-Lys > 42.5 % Minimal Reactivity Low reactivity Moderate reactivity High reactivity Non sensitiser Weak sensitiser Moderate sensitisers Strong sensitisers Gerberick GF et al. Toxicol Sci 2007, 97, 417-427
Development of alternative methods to animal testing in silico models Structure activity relationships Hapten Intracellular signaling Epidermis in chemico models hapten-protein Interactions Dendritic cells Keratinocytes in vitro models Transcriptome Phenotypical modifications Production of soluble factors Maturation Intracellular signaling Migration T TC activation Draining lymphnode in vivo models
Development of alternative methods to animal testing
KE2: KeratinoSens Mechanistic basis: address response of keratinocytes by measuring activation of the electrophile response (Nrf2-Keap1-ARE pathway) Test system: human keratinocyte-derived cell line with a luciferase gene under the control of an ARE element End-point: luciferase gene fold induction Protocol: cell exposed for 48h to 12 concentrations of test chemical; luciferase fold induction quantified by luminescence analysis Controls: cinnamic aldehyde (positive) Natsch A. Toxicol Sci 2010, 113, 284-292
Keap1 SH SH Nrf2 Proteasomal degradation Stress inducers Nrf2 Keap1 S S Target gene functions P Nrf2 Nrf2 smaf ARE Gene transcription NUCLEUS HMOX NQO1 NRF-2 Antioxidant genes Detoxication Cell Survival
KE2: KeratinoSens Prediction model: luciferase activity 1.5 fold higher at a concentration with >70% cell viability in at least 2 of 3 independent repetitions Applicability: not applicable for chemicals only reacting with lysine, non soluble substances (W/DMSO), prohaptens requiring P450 activation Natsch A. Toxicol Sci 2010, 113, 284-292
KE3: Human Cell Line Activation Test (h-clat) Mechanistic basis: address response of dendritic cells (DC) by measuring modulation of costimulatory and adhesion molecules Test system: human monocytic leukemia cell line (THP-1) End-point: relative fluorescence intensity (RFI) of CD86 and CD54 and cytotoxicity Protocol: Cells exposed for 24h to 8 concentrations of test chemical; RFI of CD86 and CD54 compared to vehicle controls quantified by flow cytometry Controls: DNCB (positive) Ashikaga, T. et al. Toxicol in Vitro 2006, 20, 767-773
KE3: Human Cell Line Activation Test (h-clat) Prediction model: A chemical is rated positive if the RFI of CD86 is 150% and/or if the RFI of CD54 is 200% at any tested dose ( 50% of cell viability) in at least 2 independent repetitions Applicability: not applicable for chemicals with logk ow greater than 3.5; prohaptens requiring metabolic activation Ashikaga, T. et al. Toxicol in Vitro 2006, 20, 767-773
Compared predictive capacity? DPRA KeratinoSens h-clat Data set size N = 82 N = 145 N = 53 Compared to LLNA LLNA Human Accuracy 89.0% 77.0% 83.0% Sensitivity 88.0% 79.0% 85.0% Specificity 90.0% 72.0% 76.9%
Major pitfalls Solubility / Toxicity Complex mixtures
Major pitfalls Activation (Pre- and prohaptens) Detoxication
Potency prediction? Potential to access Not validated
Integrated Approaches to Testing and Assessment (IATA) How to combine the results of in silico, in chemico and in vitro tests
Integrated Approaches to Testing and Assessment (IATA) Nukada et al. Toxicol in Vitro 2013, 27, 609-618 Bauch et al. Regul Toxicol Pharmacol 2012, 63, 489-504 Van der Veen et al. Regul Toxicol Pharmacol 2014, 69, 371-379
Conclusion Alternative methods developed have a potential to approach potency However they have been validated only for danger prediction IATA s are still under development for danger prediction and need to be validated All categories of sensitizers are not yet covered