prolyl hydroxylase inhibitor Following optimization chemical
Following optimization, 31 chemicals, which were considered as the skin sensitizers and non-sensitizers in LLNA studies, were studied for their potential of β-galactosidase suppression (Table 2). The detailed information on test chemicals used were given in Table 1. With optimum cell densities, E. coli prolyl hydroxylase inhibitor were co-incubated with three individual concentrations of chemicals and incubated at 37 °C for 6 h in presence or absence of pooled induced liver microsomes. Following incubation, the β-galactosidase enzyme activity and percent suppression of β-galactosidase activity by chemicals were obtained. Among 20 pre- or pro-haptens, 8 were strong sensitizers, 6 were moderate sensitizers and 6 were weak sensitizers, based on the categorization by LLNA studies (Gerberick et al., 2007; Natsch and Haupt, 2013; Troutman et al., 2011; Urbisch et al., 2016). Based on our previous study (Nepal et al., 2018a), a cut-off criterion of 17.3% was applied to all chemicals to categorize them as skin sensitizers and non-sensitizers. A decision model was developed in which a positive value in either microsome-used or -unused group would be considered as a sensitizer. All sensitizers except (R)-(+)-limonene showed positive results in either microsome-used or -unused groups. (R)-(+)-limonene, a major component of citrus fruit peels, which was considered as a weak sensitizer in LLNA studies was only predicted as a false negative in the current study when the cut-off percent of 17.3 was applied at 0.6 mM test chemical, as used in our previous study (Nepal et al., 2018a). Some chemicals, such as propyl gallate and lauryl gallate only suppressed the β-galactosidase activity when no microsome was added. Because they showed positivity in microsome-unused groups, propyl gallate and lauryl gallate were also considered as sensitizers in the present study. Similarly, among 11 non-sensitizers tested, 3 chemicals (i.e., methyl salicylate, salicylic acid, and 4-hydroxybenzoic acid) were falsely categorized as sensitizers after incubation with pooled microsomes, when a cut-off of 17.3% was applied. As shown in Table 3, if the individual prediction of either microsome-used or -used group was considered, chemicals, such as lauryl gallate and propyl gallate, were also considered as non-sensitizers and the accuracy would be lesser because of the effect of wrongly judged chemicals. However, after considering at least one positive as a sensitizer in either microsome-used or -unused group, the sensitivity was improved to 95.0%, 95.0%, and 95.0% with all the chemical concentrations tested (Table 4). Thus, a combination model was chosen as our final decision model rather than individual models (Fig. 6). With the results, the overall sensitivity, specificity and accuracy obtained were 95%, 72.7% and 87.1%, respectively, as summarized in Table 4. The results clearly showed that the incorporation of induced liver microsomes in E. coli culture system could improve the predictive capacity.
Discussion With the motive of developing in vitro alternative methods for skin sensitization, several approaches have been made, offering relatively high precision and accuracy. Nonetheless, the methods somewhat lacked the complete in vivo scenario of metabolic activation of test chemicals, by which the accurate prediction for pre- or pro-haptens would be limited. To develop a comprehensive in vitro method for skin sensitization, a test incorporated with metabolic activation system would be necessary. One of the methods for testing skin sensitization potential of chemicals recently established in our laboratory was to use a bacterial culture expressing β-galactosidase enzyme that can be induced by IPTG (Nepal et al., 2018a). During the induction of β-galactosidase by IPTG treatment, E. coli cells would be very vulnerable to the toxic insults by test chemicals due to some reasons. Most importantly, during rapid expression of certain related proteins, β-galactosidase would be affected by reactive chemicals that would be similar with the hapten interaction with proteins in the process of skin sensitization. When we tested 33 chemicals including 22 sensitizers, reasonably high predictive capacity could be obtained with relatively simple test conditions, where neither mammalian cell cultures nor in vivo studies are required (Nepal et al., 2018a). In the current study, we incorporated induced rat liver microsomes with this bacterial system to better determine skin sensitization of pre- or pro-haptens that require metabolic activation.