Anna Sonnenburg

Assessment of the sensitizing potential of textile disperse dyes and some of their metabolites by the loose-fit coculture-based sensitization assay (LCSA)

Certain textile disperse dyes are known to cause allergic reactions of the human skin. Here, we examined 8 disperse dyes and 7 products of azo-cleavage of these dyes in an in vitro assay. We used the loose-fit coculture-based sensitization assay (LCSA) of primary human keratinocytes and of allogenic dendritic cell-related cells for combined testing of the sensitizing and irritative properties of these substances. The obtained data were compared to data generated in a modified version of the local lymph node assay by our working group. Disperse Blue 1 (DB1), p-nitroaniline (pNA) and p-aminoacetanilide (AAA) showed no sensitizing potential under our experimental conditions. Disperse Blue 124 (DB124), Disperse Yellow 3 (DY3), Disperse Orange 37/76 (DO37), Disperse Blue 106 (DB106), Disperse Red 1 (DR1), 2-amino-p-cresol (ApC), Disperse Orange 3 (DO3) and 2,6-dichloro-4-nitroaniline (DCh) were categorized as extreme sensitizers. Para-phenylenediamine (pPD) was categorized as strong sensitizer, and 2-amino-5-nitrothiazole (ANT) and 2-(N-ethylanilino)-ethanol (EAE) as weak sensitizers. All dyes, except for DB1, and ApC turned out to be strong irritants. DB1, ANT and DCh showed only weak irritative potential. PPD, pNA, EAE and AAA did not show any irritative effect at the concentration range tested. These results correlate with data derived from the modified version of LLNA and human data. Therefore, the LCSA represents a suitable test system to simultaneously analyse two crucial properties of substances relevant for allergy induction.

A novel method to generate monocyte-derived dendritic cells during coculture with HaCaT facilitates detection of weak contact allergens in cosmetics

The in vitro sensitization assay LCSA (Loose-fit Coculture-based Sensitization Assay) has proved reliable for the detection of contact sensitizers in the past. However, the coculture of human monocyte-derived dendritic cells (DCs) with primary human keratinocytes (KCs) in serum-free medium is relatively complex compared to other sensitization assays which use continuous cell lines. To facilitate high-throughput screening of chemicals, we replaced KCs with the HaCaT cell line under various culture conditions. Coculture of HaCaT with peripheral blood mononuclear cells in serum-supplemented medium leads to generation of CD1a+/CD1c+ DCs after addition of GM-CSF, IL-4, and TGF-β1 (as opposed to CD1a−/CD1c− DCs which arise in the “classic” LCSA coculture). These cells resemble monocyte-derived DCs generated in monoculture, but, unlike those, they show a marked upregulation CD86 after treatment with contact allergens. All of the nine sensitizers in this study were correctly identified by CD1a+/CD1c+ DCs in coculture with HaCaT. Among the substances were weak contact allergens such as propylparaben (which is false negative in the local lymph node assay in mice) and resorcinol (which was not detected by CD1a−/CD1c− DCs in the “classic” LCSA). The level of CD86 upregulation on CD1a+/CD1c+ DCs was higher for most allergens compared to CD1a−/CD1c− DCs, thus improving the assay’s discriminatory power. Three out of four non-sensitizers were also correctly assessed by the coculture assay. A false-positive reaction to caprylic (octanoic) acid confirms earlier results that some fatty acids are able to induce CD86 on DC in vitro. In conclusion, change of the LCSA protocol led to reduction of time and cost while even increasing the assay’s sensitivity and discriminatory power.

Unsaturated compounds induce up-regulation of CD86 on dendritic cells in the in vitro sensitization assay LCSA.

Unsaturated compounds are known to cause false-positive reactions in the local lymph node assay (LLNA) but not in the guinea pig maximization test. We have tested a panel of substances (succinic acid, undecylenic acid, 1-octyn-3-ol, fumaric acid, maleic acid, linoleic acid, oleic acid, alpha-linolenic acid, squalene, and arachidonic acid) in the loose-fit coculture-based sensitization assay (LCSA) to evaluate whether unspecific activation of dendritic cells is a confounder for sensitization testing in vitro. Eight out of 10 tested substances caused significant up-regulation of CD86 on dendritic cells cocultured with keratinocytes and would have been classified as sensitizers; only succinic acid was tested negative, and squalene had to be excluded from data analysis due to poor solubility in cell culture medium. Based on human data, only undecylenic acid can be considered a true sensitizer. The true sensitizing potential of 1-octyn-3-ol is uncertain. Fumaric acid and its isomer maleic acid are not known as sensitizers, but their esters are contact allergens. A group of 18- to 20-carbon chain unsaturated fatty acids (linoleic acid, oleic acid, alpha-linolenic acid, and arachidonic acid) elicited the strongest reaction in vitro. This is possibly due to the formation of pro-inflammatory lipid mediators in the cell culture causing nonspecific activation of dendritic cells. In conclusion, both the LLNA and the LCSA seem to provide false-positive results for unsaturated fatty acids. The inclusion of T cells in dendritic cell-based in vitro sensitization assays may help to eliminate false-positive results due to nonspecific dendritic cell activation. This would lead to more accurate prediction of sensitizers, which is paramount for consumer health protection and occupational safety.

Erratum to: A novel method to generate monocyte-derived dendritic cells during coculture with HaCaT facilitates detection of weak contact allergens in cosmetics

Authors would like to correct the errors in the published article. The first sentence of the second paragraph of the materials and methods section needs to be changed to: “Dimethylsulfoxide (DMSO; Sigma-Aldrich) was used as a vehicle for DNCB, MBT, caprylic acid, and parabens.” The two EC50sens values for caprylic acid in LCSA setup B and C were accidentally interchanged. The fourth sentence of the second paragraph of the results section needs to be changed to: “Thus, EC50sens was >252 μmol/l (setup A), >219 μmol/l (setup B), and >669 μmol/l (setup C).” In the last row of Table 1, EC50sens should be “>219” in column B and “>669” in column C.