Assessment of Cutaneous Permeability of Biocides in Mixtures using QSPR Approach


The purpose of this research work was to assess the dermal permeation of biocides in metalworking fluids (MWFs) to develop predictive QSAR models and to develop an appropriate training set of chemicals to enhance the predictive ability of QSAR models for dermal permeation. Estimation of the amount of chemicals absorbed through skin plays a vital role in dermal risk assessment. Approximately 1.2 million US workers are occupationally exposed to MWFs annually. Different components of MWFs especially biocides, contribute to adverse health effects including irritant and allergic contact dermatitis as well as carcinogenesis. These adverse effects may be positively correlated to their dermal absorption and may cause systemic toxicity if absorbed in significant amount in workers involved in metalworking operations. A lack of scientific data exists regarding the dermal permeation of MWF components, particularly biocides. Therefore, the first two studies were conducted to (1) determine the dermal permeation of biocides and other chemicals (used as training set to develop Linear Solvation Energy Relationship (LSER) models) in commercial and generic MWFs; and (2) develop a LSER model for predicting dermal permeation of other biocides, not used in these studies. Dermal permeation was evaluated in dermatomed porcine skin by utilizing a flow through diffusion cell system. Chemical analysis was performed by employing gas chromatography with a solid phase micro-extraction technique and ultra performance liquid chromatography with a solid phase extraction technique. LSER models, which are a subset of quantitative structure activity relationship models, were constructed by multiple linear regression analysis with permeability coefficient as the response variable and solvatochromic descriptors as the predictor variables. The LSER model is useful to quantitatively measure the difference in interaction between the two phases (skin and vehicle) as well as a predictive tool. Since the training set used to develop a LSER model was not optimally diverse in terms of structure and chemical space, the third study focused on developing a training set of chemicals representing a wider chemical space (in terms of descriptor values) using a best possible chemical selection method. The results from the first two studies demonstrated that (1) the dermal permeation of biocides as well as other chemicals was highest in aqueous solution followed by synthetic, semi-synthetic and soluble oil type of MWFs; (2) addition of water to MWFs for dilution increased dermal permeation; (3) the LSER model adequately predicted the dermal permeability of biocides in MWFs and also shed light on the chemical interactions resulting in reduced permeability. An optimal and less subjective method (uniform coverage design) to select chemicals representing a wider chemical space was identified in the third study. The LSER model based on the new selected training set of chemicals performed statistically better over the LSER model based on the training set used in the previous study. In summary, the aforementioned results demonstrated that there is a difference in the absorption profile of chemicals among the type of MWFs and dilution of MWFs with water increases the dermal permeation of chemicals; the LSER model can be useful to explain the change in vehicle solvatochromic properties upon addition of water as well as can be an effective prediction model for dermal permeation of chemicals in mixtures; finally, a structurally diverse training set of chemicals representing a wider chemical space is required to improve the predictive capability of a model. All of these results will augment the dermal risk assessment of the chemicals in mixtures and contribute to the improvement of computational predictive models.



LSER, QSAR, modeling, dermal permeation, biocides, metalworking fluids, diverse chemical selection, occupational exposure





Comparative Biomedical Sciences