Permeability of 16 Straight- and Branched-Chain Parabens Using the Caco-2 Assay

The in vitro intestinal permeability of straight- and branched-chain parabens has not been extensively investigated. Sixteen parabens were tested in the Caco-2 assay. Passive diffusion was measured using PAMPA. The transport of the MCT1 substrate, p-coumaric acid, as well as propylparaben and isopropylparaben, was investigated. For straight-chain parabens, P_{app A-B} and P_{app B-A} decreased with increasing chain length and LogP. P_{app B-A} values were similar to PAMPA permeability (P_{app PAMPA}), indicating passive diffusion. Losses in mass balance were due to non-specific binding, accumulation in the cells, and/or hydrolysis to 4-hydroxybenzoic acid (4-HBA). The extent of hydrolysis of straight-chain parabens was inversely proportional to their LogP, suggesting they are carboxylesterase-1 (CES1) substrates. For C1-C5 straight-chain parabens, P_{app A-B} was higher than P_{app B-A}, indicating vectoral permeability. Transport of parent propylparaben was passive and pH-independent, but 4-HBA formed was actively transported out of the cells, which was pH-dependent. This indicated the involvement of apical MCT1 transporters (their presence was confirmed using p-coumaric acid). Kinetics measurements suggested that efflux of 4-HBA is predominantly via the basolateral membrane. Branched-chain parabens with good passive diffusion were poor CES1 substrates but may be transported via processes other than 4-HBA MCT1 efflux. In conclusion, P_app values for parabens are best calculated using parent chemical and 4-HBA. The extent of vectoral permeability of straight-chain but not branched-chain parabens is correlated to the extent of hydrolysis and the concentration-dependent contribution of passive vs. active efflux of parent and 4-HBA, respectively.