DescriptionAllergic contact dermatitis is an inflammatory skin disease that impacts 15-20% of the general population. The prevalence of potentially sensitizing agents necessitates screening methods for chemical risk assessment. Variability, cost, ethical concerns, and the recent ban of animal testing in the European Union introduce a pressing need to develop alternative methods to animal testing to screen for the sensitization potential of chemicals. Many current screening approaches are limited in their ability to predict pre- and pro-hapten sensitizers, which require abiotic or metabolic conversion prior to inducing sensitization, and sensitizer potency. We previously developed an in vitro co-culture system of MUTZ-3-derived Langerhans cells, HaCaT keratinocytes, and primary dermal fibroblasts to mimic the in vivo cellular and metabolic environments of skin sensitization. We expanded the chemical test panel to include a variety of non-sensitizers, haptens, pre-haptens, and pro-haptens of all potencies and compared the co- culture system’s performance to MUTZ-3 Langerhans cells alone by measuring CXCL8 secretion. The secretome of both cultures were also evaluated for 27 cytokines, chemokines, and growth factors. A support vector machine was used to identify the most predictive signature of sensitization for each culture system and classification trees were used to identify statistical thresholds to predict sensitizer potency by CXCL8 secretion. The support vector machine computed prediction accuracy of 87% for the MUTZ-3 mono-culture system using the top 12 ranked biomarkers while predicting accuracy of 91% for the co-culture system using the top 3 (IL-8, MIP-1β, and GM-CSF). The classification trees demonstrated 83% accuracy for potency prediction by the co-culture system and 73% accuracy for MUTZ-3 cells alone. Overall, the presence of keratinocytes and fibroblasts enhanced the system’s ability to detect pre- and pro- haptens, resulting in higher accuracy scores for the co-culture system. To further reduce the use of animal derivatives, other predictive metrics are being explored. Further studies using the cells of our co-culture demonstrate that the Langerhans cells are requisite for CXCL8 secretion, and therefore RNA expression in sensitized Langerhans cells from co-culture will be assessed to identify additional predictive metrics. Mitochondrial markers of cellular stress due to sensitization are also being pursued in this system. As such, this co-culture in vitro assay presents a promising alternative method to animal testing for screening and classifying potential skin sensitizers and offers the possibility of multi-metric analysis for enhanced screening capabilities and mechanistic studies of allergic contact dermatitis.